Dev Summit 2026: Full Schedule

10:00 CEST

Causality at scale: From statistical models to AI agents at Decathlon

Tuesday May 26, 2026 10:00 - 10:30 CEST

In a complex omnichannel environment like Decathlon’s, distinguishing between correlation and true incremental impact is the ultimate challenge. Traditional A/B testing often falls short when physical stores, local events, and global trends collide.

This session dives into our journey of scaling measurement through Randomized-Controlled Trials (RCTs) and Geo-experiments. We will explore how we utilize techniques to quantify the real value of business use cases, like testing the price elasticity of the products.

However, the real breakthrough lies in how we think about the evolution of this process. We will showcase our vision of Agentic AI: moving from manual statistical workflows to autonomous AI Agents that can help on the experiment design, select optimal control markets, and translate them into actionable business insights.

It's a great opportunity to see how we are democratizing rigorous data science to ensure every decision at Decathlon is backed by the "Truth of Causality."

Speakers

Sergio Benito

Decathlon

Passionate about sports and its benefits as well as helping people make better decisions through Data and Artificial Intelligence. I'm working at Decathlon for almost 10 years coming from the retail and thanks to that experience I joined at the Data & AI Spanish team, where I try... Read More →

Tuesday May 26, 2026 10:00 - 10:30 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

10:00 CEST

GOOGLE X ADEO - How to reduce BigQuery costs? democratizing best practices at scale.

Tuesday May 26, 2026 10:00 - 10:30 CEST

🤖 DATA/AI ARENA

Joint session between Adeo's BigQuery Finops team and Google Cloud specialists to highlight how Adeo monitors and proposes optimisations to its business units, and how it impacts the performance and costs at all levels.

Speakers

Olivia Paris

Adeo Services

Léo Negre

Data Analytics Specialist, GOOGLE CLOUD

Tuesday May 26, 2026 10:00 - 10:30 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

10:00 CEST

NASDAQ - Problem First: How to identify and build high-value AI in complex systems?

Tuesday May 26, 2026 10:00 - 10:45 CEST

1 - MAIN STAGE

AI is moving fast - and so are the decisions that come with it. For engineering teams working in complex, real-world systems, the challenge isn’t access to AI technology. It’s knowing where to start, what to build, and how to make the right choices across an increasingly sophisticated stack.
This talk shares practical approaches to identifying high-value AI use cases, drawn from experience designing and deploying production machine learning systems inside one of the world’s most demanding environments - global financial market surveillance across more than 50 exchanges and 15 regulators worldwide.
A core theme is the relationship between problem definition and technology selection. Before choosing a tool, understanding the process, the inputs and outputs, and what genuinely better looks like tends to produce stronger outcomes - and clearer architecture decisions.
The session also explores how classical ML, generative AI, and agentic systems can work together effectively. Rather than viewing these as competing approaches, the talk offers a framework for thinking about each as a distinct capability - and for combining them deliberately to build systems that are greater than the sum of their parts.
Attendees will leave with practical frameworks they can apply to their own environments, a clearer mental model for navigating the modern AI stack, and fresh perspectives on how some of the most complex AI systems in the world are being designed and deployed today.

Speakers

Sarah Bradley

Director of Data Science & AI, NASDAQ

Sarah Bradley is Director of Data Science & AI at Nasdaq, where she leads the development and deployment of machine learning, generative AI, and agentic AI capabilities across Nasdaq Market Surveillance - a platform used by more than 65 exchanges and regulators globally to detect... Read More →

Tuesday May 26, 2026 10:00 - 10:45 CEST
1 - MAIN STAGE

Data / AI, Conference

Audience Intermediate

10:45 CEST

PaintMyWall: From AI hackathon to production.

Tuesday May 26, 2026 10:45 - 11:15 CEST

🤖 DATA/AI ARENA

Generative AI opens exciting opportunities to reinvent retail experiences. However, transforming a promising demo into a reliable product feature is far more complex than it first appears.

In this talk we will present how we moved from an AI demo built for a hackathon to a full fledged AI feature pushed to the millions of users of our mobile application.
Integrating image generation models into an existing application presents unique architectural challenges. How do you ensure the model quality despite the constraints of a mobile application? How do you balance costs with massive user scale?

We will trace the PaintMyWall journey from early AI experiments to introducing mobile capabilities, allowing customers to test colors on their walls with a simple photo, before taking a look at the opportunities this technology unlocks.

Speakers

Damien Toulouse

Software Engineer, Ext - ADEO Services

I'm Damien Toulouse, a Software Engineer specialized in mobile app development since 2018, i've spent the last few years working on the Customer Mobile App first as a native Android developer then as a Flutter Developer.

Baptiste Thiriet

Back-End Developer, Ext - ADEO Services

Pacome Morisot

ADEO Services

Tuesday May 26, 2026 10:45 - 11:15 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

11:15 CEST

From "baton passes" to building blocks: Making data work as a team sport.

Tuesday May 26, 2026 11:15 - 12:00 CEST

2 - GRAND PAVILION

What do sports, home improvement, and modern AI have in common? None of them work well when handoffs are messy, foundations are shaky, or every new initiative starts by tearing down what already exists. This keynote explores how data products and data contracts help organizations create clearer ownership, stronger collaboration, and data that is truly ready for products, analytics, and AI through a practical, frugal, and brownfield-friendly approach.

In many organizations, the challenge is no longer collecting quality data. It is making that data trustworthy enough, with sufficient context, to be reused across teams, embedded in products, and pushed all the way into AI systems that actually reach production.

In this keynote, Jean-Georges “jgp” Perrin explores how data products and data contracts help organizations build stronger foundations for collaboration, delivery, and AI success. Using simple analogies from sport and construction, he shows why better outcomes depend on clean handoffs, explicit expectations, and assets designed to be reused rather than rediscovered every quarter.

The session also introduces Bitol, the Linux Foundation project advancing open standards for modern data engineering, including the Open Data Contract Standard (ODCS) and the Open Data Product Standard (ODPS). These standards are not abstract theory: they reflect a growing, proven practice in which data contracts and data products help teams reduce ambiguity, improve quality, and move AI initiatives from promising prototypes to production with greater confidence.

Most importantly, this is not a rip-and-replace story. It is a frugal, brownfield approach for real organizations: improve what exists, strengthen trust step by step, and create measurable value without wasting time, effort, or money. Built the way strong teams in sport, on worksites, and in the Nord often succeed best: with coordination, clarity, and something sturdy enough to last.

Speakers

Jean-Georges Perrin

Senior Product Manager, Actian

Jean-Georges “jgp” Perrin is a data passionate. As chair of the LF’s Bitol project, Jean-Georges leads efforts to establish global standards for data, including ODCS. He is the author of Implementing Data Mesh (O’Reilly) and Spark in Action 2e (Manning). He is currently... Read More →

Tuesday May 26, 2026 11:15 - 12:00 CEST
2 - GRAND PAVILION

Data / AI, Conference

Audience Beginner

11:30 CEST

Transforming market expansion at Leroy Merlin Brazil: From descriptive geography to AI-Driven geo-intelligence.

Tuesday May 26, 2026 11:30 - 12:00 CEST

🤖 DATA/AI ARENA

The Geo-Intelligence initiative represents a strategic shift in how market expansion opportunities are identified and prioritized. By leveraging the ArcGIS platform (a leading Geographic Information System - GIS, used to analyze, visualize, and manage spatial data), we developed an AI-driven, state-of-the-art geospatial framework that integrates spatial data, market indicators, and internal sales performance to support data-driven geographic decision-making.

The analytical architecture combines state-of-the-art statistical modeling techniques, geospatial data enrichment, and automated ZIP code geolocation, enabling accurate estimation of market potential and current market penetration at a granular region level.

This approach also supports marketing and commercial strategy, helping teams identify high-value territories, target campaigns more effectively, and allocate sales resources based on geographic opportunity.

In addition, the framework incorporates state-of-the-art predictive models for sales forecasting, allowing the organization to estimate potential revenue and prioritize expansion areas before investments are made.

Integrated within the ArcGIS ecosystem, this initiative transforms geographic analysis into a scalable decision-support capability, enabling faster, data-driven expansion strategies and improved revenue growth planning.

Speakers

Eliana Junko Ishizuka

Leroy Merlin Brazil

Eliana has been working at the Leroy Merlin Brazil for 11 years in the Market Research area within the Expansion division. She holds a degree in Geography from the University of São Paulo (USP) and has strong experience in geospatial data analysis applied to business. Throughout... Read More →

André Rabetti Alves

Data Scientist, Leroy Merlin Brazil

André Rabetti is a Data Scientist with over 4 years of experience in data analysis, market intelligence, and machine learning, working on transforming data into strategic business decisions. He is currently part of the data team at Leroy Merlin Brazil, where he works directly with... Read More →

Tuesday May 26, 2026 11:30 - 12:00 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

11:45 CEST

Ranking is a decision: Engineering a data driven truth system for AI-powered omni-channel search.

Tuesday May 26, 2026 11:45 - 12:15 CEST

3 - LAB

**Every time a search engine displays a product, it makes a decision**. Not just a sort order, but a business decision that impacts customer experience, revenue, and fairness between competing products and channels. But what actually defines a good decision?
**A ranking system is effectively a judge: every query is a trial, every product is a candidate, and user behavior becomes the evidence**. The challenge is that this evidence is incomplete, biased, and distributed across channels. Clicks measure attention but not intent, while conversions capture purchases but miss delayed or offline behavior.
Search ranking must also balance competing objectives. The products people click are not always the ones they buy. Visually appealing items attract curiosity and traffic, while the products that generate revenue may appear less attractive at first glance. Optimizing purely for purchases can also favor high volume, low cost items and erode overall business value.
Time adds another layer of complexity. Some products perform consistently year after year, while others spike during seasonal moments or emerging trends. Ranking systems must decide whether to trust long term reliability or react to short term signals.
At the same time, behavioral signals are distorted by position bias, sparse data, and delayed attribution across channels. **In an omnichannel environment where customers research online and purchase later in store, the ROPO (Research Online Purchase Offline) effect makes naive metrics misleading**.
To address this challenge, the Search and Publication team developed a decision pipeline that industrializes the production of implicit judgments and acts as a **Truth Provider for AI systems**. The system transforms noisy behavioral signals into stable implicit judgments used to evaluate ranking systems, guide experimentation, train machine learning models, and provide reliable ground truth signals used by multiple internal products.
Built on top of BigQuery and dbt, the pipeline processes hundreds of millions of search sessions and several millions of products every day. It integrates multi channel evidence, bias correction, Bayesian priors, uncertainty estimation, temporal weighting, and composite indicators balancing discovery and purchase signals.
Attendees will gain a practical understanding of how to transform noisy behavioral data into reliable implicit judgments, and how to choose the right signals when discovery, conversion, and long term reliability pull ranking decisions in different directions.
**Ranking is not just an algorithm problem. It is a decision problem**. Every decision needs an honest definition of what good actually means.

Speakers

Mathieu Deleu

Lead Data Science, Ext - ADEO Services

Mathieu Deleu, Lead Data Science

Tuesday May 26, 2026 11:45 - 12:15 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

12:15 CEST

Agentic commerce 2026: When AI becomes your most important customer

Tuesday May 26, 2026 12:15 - 12:45 CEST

🤖 DATA/AI ARENA

The e-commerce landscape is entering a long-term transition. While humans still hold the credit card, AI agents are increasingly taking over the discovery, comparison, and decision-making phases. With projections suggesting a significant portion of commerce will be agent-mediated by 2030, the shift for retailers starts now.

In this dual-voice session, a **Manager from the world of the Data** and a **Staff Software Engineer** bridge their perspectives to provide a pragmatic state of the union on Agentic Commerce. We will move beyond the hype to look at the incremental reality of this hybrid era:
* **Strategic Perspective**: How to remain relevant when an agent is filtering your catalog? We’ll discuss the shift in brand loyalty and the challenge of serving both human shoppers and their digital assistants.
* **Technical Foundations**: Preparing your stack for machine-to-machine interactions. We will cover the evolution of MACH architectures, the importance of goal-oriented APIs, and the role of standardized protocols like *MCP*.
* **Practical Scenarios**: From automated replenishment of consumables to "Intent-based" shopping, where agents orchestrate complex purchases across multiple categories.
Join us to explore the architectural patterns and strategic shifts that will bridge the gap between today’s web and tomorrow’s agentic ecosystem.

Speakers

Cyril Gambis

Decathlon

After completing my Master's degree in Computer Science & Applied Mathematics at a Grenoble engineering school (2002), I developed a strong passion for emerging technologies and systems architecture.

My career has taken me through a diverse range of environments, from IT consulting firms and investment banks to software companies and startups. I have now joined Decathlon to help drive its technological advancement. My core specialty lies in distributed systems and microservices... Read More →

Tuesday May 26, 2026 12:15 - 12:45 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Beginner

14:00 CEST

From "Intuition" to data-driven precision: Automating store safety stocks at scale

Tuesday May 26, 2026 14:00 - 14:30 CEST

🤖 DATA/AI ARENA

In the world of global retail, balancing product availability with inventory cost is a high-stakes game of precision. For years, managing "Safety Stock"—the buffer inventory held to mitigate demand and supply uncertainties—at Decathlon relied heavily on empirical rules, based on fixed numbers of coverage days regardless of the product, inducing high-level coverage values. While functional, this "one-size-fits-most" approach lacked the granularity needed to truly optimize a complex network of thousands of stores and unique items.

This session details the technical evolution of our Safety Stock solution, moving from these empirical rules to an automated, industrial-scale statistical engine. We will present a deep dive into how we built a system that computes dynamic safety stocks at the Item x Store x Day level.

Key technical pillars of the session include:
- The Statistical Engine: Moving beyond the baseline to formulas integrating forecast error variability, and lead time variability to meet specific target service levels.
- The Architecture: An overview of our data pipeline, utilizing dbt for weekly computations based on Decathlon's datalake and Talend flows to push granular safety stock quantities directly into SAP F&R.
- Backtesting & Impact: A transparent review of our pilot results across pilot product families, demonstrating a reduction in Days Sales of Inventory (DSI) of a few days without sacrificing availability.
- Product Segmentation: How we use ABC-XYZ classification to automatically tailor service level targets based on both business targets and demand predictability.

Join us to learn how we bridged the gap between supply chain theory and real-world data science and analytics engineering, transforming inventory management from a "best guess" into a precise, automated competitive advantage.

Speakers

Vianney Taquet

Data & AI Manager, Decathlon

Data & AI manager at Decathlon since September 2022 on demand forecasting and inventory management within Decathlon Digital's supply chain department.

Tuesday May 26, 2026 14:00 - 14:30 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

14:00 CEST

GOOGLE - A semantic and contextual Data Platform

Tuesday May 26, 2026 14:00 - 14:45 CEST

3 - LAB

Agents need to be grounded context and semantic to operate properly. Focus on Google Cloud's vision and trajectory on semantic layers and data governance

Speakers

Johan Picard

Data Analytics Leader, GOOGLE

I am leading Google Cloud's Data Analytics Technical Practice Team in France and, as such, am working closely with most of our FR customers in defining both their data platforms and strategies !

Tuesday May 26, 2026 14:00 - 14:45 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

14:00 CEST

NASDAQ - AI in the wild: Survival lessons from real-world model development

Tuesday May 26, 2026 14:00 - 14:45 CEST

2 - GRAND PAVILION

AI looks clean in tutorials. In production, it rarely is. This workshop offers a practitioner’s walkthrough of real-world ML development - exploring what the process actually looks like when you’re building detection systems for real clients, with real data, under real constraints.
Drawing on experience developing and deploying machine learning systems inside global financial market surveillance, the session is structured around three transferable lessons: how to translate a client’s business problem into a well-formed ML problem; how to stay flexible and recalibrate when early assumptions don’t hold; and how to navigate the statistical realities of imbalanced, dependent, and noisy real-world data.
Each lesson is grounded in real project experience - honest about the complexity, and focused on the thinking and decision-making that determines whether a model earns its place in production. Attendees will leave with practical insight they can apply to their own projects, and a clearer picture of what separates ML that ships from ML that stalls.

Speakers

Sarah Bradley

Director of Data Science & AI, NASDAQ

Tuesday May 26, 2026 14:00 - 14:45 CEST
2 - GRAND PAVILION

Data / AI, Talk

Audience Advanced

14:45 CEST

Beyond the cookie: Building a zero-knowledge ad ecosystem with fully homomorphic encryption (FHE)

Tuesday May 26, 2026 14:45 - 15:15 CEST

🤖 DATA/AI ARENA

The "Post-Cookie" era has left the digital ecosystem at a high-stakes crossroads: sacrifice measurement and user experience, or sacrifice privacy. While current workarounds like "Clean Rooms" and "Contextual Targeting" offer temporary relief, they fail to bridge the gap between **total user anonymity** and **hyper-relevant engagement**.

This talk introduces **Fully Homomorphic Encryption (FHE)** as the ultimate architectural pivot—a cryptographic framework that allows for both precise attribution and real-time website personalization without ever decrypting user data. We will explore how FHE enables "Blind Personalization," where a server can rank products and serve tailored content to a user it doesn't "know," and how it solves the attribution loop with mathematical certainty.

By moving past the "FHE is too slow" myth, we will demonstrate how 2026’s hardware acceleration and edge computing are making **Zero-Knowledge Advertising**—from the first click to the final purchase—a production-ready reality.

Speakers

Francois Serra

ML Engineer / MLOps / Versatilist, ADEO Services

François Serra is a seasoned Machine Learning Engineer at ADEO. With over 20 years of expertise in software engineering and architecture, he has seamlessly transitioned his skills into the AI/ML domain, establishing a robust track record. François has successfully built and deployed... Read More →

Tuesday May 26, 2026 14:45 - 15:15 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Advanced

15:00 CEST

A design process for dashboards

Tuesday May 26, 2026 15:00 - 15:30 CEST

3 - LAB

Today, companies are producing more dashboards than ever before. Yet, many organizations find themselves with what I call "dashboard debt": dozens of screens, rarely used, and seldom linked to actual decisions.

We often hear that companies want to be data-driven. But in reality, many are becoming dashboard-driven instead. In this talk, I will defend a simple idea: stop delivering charts, and start delivering decisions.

The problem rarely lies with the charts themselves. It comes from what happens beforehand: a lack of discovery, requirements that are accepted too quickly, and solutions that are decided upon before the problem is even understood. When everything becomes a KPI, nothing is truly "key" anymore. And when an organization has 3 to 10 dashboards per employee, it's generally not a very positive sign of data debt management.

I will show how data teams can shift from a logic of producing dashboards to a logic of creating data products designed around real needs, thanks to a structured process of discovery, prioritization, and design. The goal is clear: fewer dashboards, but much more value.

Speakers

Aurélien Vautier

Dataviz Clarity

Aurélien Vautier est fondateur de Dataviz Clarity, une initiative dédiée à améliorer la manière dont les entreprises conçoivent et utilisent leurs produits data.

Depuis plus de dix ans, il travaille avec des équipes data et analytics pour transformer des dashboards souvent conçus comme de simples livrables en véritables outils d'aide à la décision. Son travail se concentre sur l'introduction de méthodes de design, de discovery et de... Read More →

Tuesday May 26, 2026 15:00 - 15:30 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

15:00 CEST

GOOGLE DEEPMIND - AI for Systems for AI: There and Back Again

Tuesday May 26, 2026 15:00 - 15:45 CEST

1 - MAIN STAGE

Lessons from Deploying AI in Google's Production Infrastructure, from the early days of AI for systems design and optimization to agentic AI workflows. We will cover examples from optimization research (logistics, warehouse design), distributed systems, code optimization, and AI-first software-engineering practices. We will also zoom in on deployment and adoption challenges.

Speakers

Albert Cohen

Research Scientist, Google Deepmind

Tuesday May 26, 2026 15:00 - 15:45 CEST
1 - MAIN STAGE

Data / AI, Conference

15:30 CEST

L'OREAL X GOOGLE - PowerBI on Bigquery at scale

Tuesday May 26, 2026 15:30 - 16:00 CEST

🤖 DATA/AI ARENA

Speakers

Antoine Castex

Group Data & A.I Architect, L'OREAL

Keep it simple
Serverless is More
Eat what you cook
May the Cloud be with you

BUREL Mathieu

Strategic Program - BI & Analytics Manager, L'Oréal

With 20 years of experience in Data Intelligence, I help global organizations elevate their analytics ecosystems by bridging technology, business context, and user experience. From consulting to corporate, from reporting stacks to product-led BI, my journey is rooted in execution... Read More →

Tuesday May 26, 2026 15:30 - 16:00 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

15:45 CEST

Beyond the chatbot: Engineering a multi-agentic lego-like ecosystem

Tuesday May 26, 2026 15:45 - 16:15 CEST

3 - LAB

The era of monolithic LLM assistants is giving way to a more powerful, resilient reality: Agentic Ecosystems. In this session, we will explore how to transition from isolated AI tools to a modular, 'Lego-like' architecture of autonomous agents that reason, plan, and collaborate.

We will dive deep into the technical 'glue' that enables this synergy, focusing on ADK for orchestration, and the critical role of protocols like MCP (Model Context Protocol) and A2A (Agent-to-Agent) for seamless communication. You will learn how we differentiate between specialized sub-agents—focused on niche expertise like data or validation—and the 'Coordinator' agent that acts as the system’s brain to plan and validate results.

Using INES (Our Digital Instore Assistant) as our primary case study, we will demonstrate how this ecosystem provides high-value operational efficiency by simplifying access to information for store teams. We’ll share our lessons on how to work on an independent agentic ecosystem.

Speakers

Ruben Sastre

Decathlon

Daniel Rozputyñski

Decathlon

Data Engineer & AI Engineer with 9 years of experience in Analytics & AI world, passionate about sports and technology.

Tuesday May 26, 2026 15:45 - 16:15 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

16:15 CEST

Beyond the hype: 3 Pentests and 2 GenAI services later, an LLM security survival guide

Tuesday May 26, 2026 16:15 - 16:45 CEST

🤖 DATA/AI ARENA

Generative AI is no longer just a "Lab" project, it’s in production. But when you move from a playground to an industrial environment like Decathlon, the stakes change. Suddenly, "Prompt Injection" isn't a fun riddle; it's a potential data breach or a compliance nightmare.

In this session, we (PIC Data Team & Cybersecurity Team) share our raw, unfiltered journey of securing two major GenAI applications: a global translation engine and a content compliance engine.

We didn't just read the OWASP Top 10 for LLMs—we lived it through three rigorous professional pentests. We will walk you through the actual vulnerabilities the ethical hackers found and, more importantly, the concrete defense layers we built in our GenAI services to neutralize them.

Speakers :
Pierre-Yvan Devos (Data & AI Manager)
Karl Mongosso (Senior Machine Learning Engineer)
Philippe Lasek (Information Security Officer)

Speakers

Philippe Lasek

Security Officer, Decathlon

Security Officer since 2018, first in IAM & People Business Unit and now for the Conception Business Unit.
Previously : Google Admin, Engineering Manager.
Deployment of DMARC and BIMI on our main domain mail.

Karl Mongosso

Senior Machine Learning Engineer, Decathlon

Tuesday May 26, 2026 16:15 - 16:45 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

16:30 CEST

IA11Y: What's next for accessibility with AI?

Tuesday May 26, 2026 16:30 - 17:00 CEST

3 - LAB

2025 was supposed to be a major year for accessibility with the enforcement of the European Accessibility Act. But the AI tsunami came in and stole the show! The future of technology seems inextricable of artificial intelligence. However is AI and Accessibility really a good mix ?

People with disabilities are already using AI tools in their daily lives: to navigate the web, understand content, or overcome barriers that were previously difficult to address. At the same time, AI-generated interfaces and AI-assisted development introduce new challenges: inaccessible tools, biased outputs, unreliable alternative text, or code that looks correct but fails basic accessibility requirements.

This talk looks at the current intersection between accessibility and AI. Through practical examples and a case study of a “vibe-coded” website, we will explore where AI can genuinely help improve accessibility and where it still falls short. The goal is not to present AI as a solution to accessibility, but to understand how it can support the work of designers and developers without replacing the human perspective that accessibility ultimately depends on.

Speakers

Pierre Leclaire

Ext - ADEO Services

**Version courte pour la conférence :**

Je suis développeur frontend, passionné par l’accessibilité web. J’ai découvert ce sujet lors d’un DevLille il y a quelques années, et il est rapidement devenu central dans mon travail. Aujourd’hui, je conçois et anime des formations sur l’accessibilité chez SFEIR... Read More →

Tuesday May 26, 2026 16:30 - 17:00 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Beginner

17:00 CEST

Pricing decisions: From dashboards to a production-grade AI assistant

Tuesday May 26, 2026 17:00 - 17:45 CEST

🤖 DATA/AI ARENA

At Decathlon, our Pricing Decision Makers were suffering from *infobesity*. We solved that by changing their UI from 12 complex Tableau dashboards into a **frictionless, natural-language AI assistant** integrated directly into Google Chat.

But taking a **Generative AI agent** from a **shiny POC** to a highly accurate robust **production** product is paved with technical hurdles.
In this talk, we will **open the hood** and share our engineering learnings.

LLMOps, monitoring, evaluation, costs ... Join us to discover what it actually takes to build and **run GenAI at scale**!

Speakers

Nicolas Gorrity

Decathlon

Nicolas GORRITY - Senior ML Engineer

- 3 ans d'expérience en Data Science dans l'automobile puis dans le Retail.
- 3 ans d'expérience en ML Engineering dans le Retail, dont 2 ans chez Decathlon Digital.

Tuesday May 26, 2026 17:00 - 17:45 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

17:15 CEST

AI-based virtual coaching for sales advisors to master product knowledge

Tuesday May 26, 2026 17:15 - 17:45 CEST

3 - LAB

In an era where product complexity is rising and customer expectations are higher than ever, traditional training often falls short of the "real-world" pressure of the sales floor. Explore how AI Virtual coaching is bridging the gap between technical product knowledge and expert consultation.

This talk introduces an innovative Customer AI-Simulator for sales advisors co-built with sellers experts from LMIT and learning teams, and approved by UX. By leveraging Speech-to-Text and AI Dialogue, we have created a safe, interactive environment where sales advisors can practice complex scenarios, such as the technical "Shower Enclosure" family, with virtual customers to master their knowledge and be able to traslate produt characteristics in benefits for the customer.

Longer description:
The Problem that solution solves
- New advisors struggle to internalize the Leroy Merlin sales model.
- Senior advisors know the specs, but often struggle to translate them into customer benefits.
- Classroom training is disconnected from the daily reality of the store floor.

The Solution: The AI Virtual Coach MVP
- We’ve built a Customer AI-Simulator. This isn't just a chatbot; it’s a high-fidelity 'flight simulator' for retail. - a verbatim from a seller that co-designed this coach.
- Real-time Interaction: Using Speech-to-Text and AI Dialogue, advisors engage in a natural verbal flow with a virtual customer.
- Realistic Scenarios: From "budget-conscious" to "aesthetic-oriented," the AI mimics real human behaviors and objections.
- Immediate Feedback Loop: After the 'sale,' the AI provides an instant debrief based on LMIT standards—identifying strengths, missed opportunities, and discovery depth.

Why it’s a Game Changer (The Value)
- Risk-Free Practice: Advisors can fail, learn, and iterate in a safe environment.
- Standardization: Aligns every store to the same high-quality LM selling model (M2, M3, M4).
- Data-Driven Growth: We finally have visibility into how our teams sell, not just what they know.

Speakers

Albina Muzyka

Business Domain leader New ways of working, ADEO Services

Albina Muzyka is a leader in Digital HR and Employee Experience Transformation at Adeo Services, with 15 years of international experience across the ADEO ecosystem. Having spent nearly a decade in HR at Leroy Merlin Ukraine before moving into global digital transformation, Albina... Read More →

Tuesday May 26, 2026 17:15 - 17:45 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Intermediate

17:45 CEST

From User Needs and Behavior to Data Modeling: Reducing FinOps Costs with Smarter Modeling

Tuesday May 26, 2026 17:45 - 18:15 CEST

🧩 PM/UX ARENA

Data models are often designed around data availability and development convenience, but not always around how users actually consume data. As analytical products scale, this misalignment can significantly increase infrastructure costs.
In this talk, I will share the transformation of the Seller Deep Dive Dashboard, which at one point became one of the most expensive dashboards to operate due to its data model design.
The initial architecture relied on a fully denormalized table, where dimensions were merged directly into the fact table powering the dashboard. While this simplified early development, the model became increasingly inefficient as data volume grew and slowly changing dimensions expanded the dataset.
By analyzing which KPIs users actually queried and how frequently they were used, we redesigned the model around a usage-driven star schema.
The new architecture separates metrics into two fact tables:

A core fact table containing roughly 80% of the most frequently used KPIs, representing only 20% of the total data weight

A secondary fact table containing the remaining less frequently used but heavier metrics, queried only when required

This change dramatically reduced the amount of data scanned by the majority of queries.
As a result, the dashboard evolved from being one of the most expensive dashboards to run, costing roughly €3500 over two weeks, to becoming one of the most efficient, reducing costs to approximately €350 over the same period—a 90% reduction in FinOps costs while maintaining full analytical capability.

Speakers

Juan Jose Medina Martinez

Data Analyst, ADEO Services

Juan José Medina - Data Analysts - Adeo Marketplace

Tuesday May 26, 2026 17:45 - 18:15 CEST
🧩 PM/UX ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

17:45 CEST

Last Mile Delivery Intelligence: AI-Powered Road Accessibility Analysis for Fleet Planning

Tuesday May 26, 2026 17:45 - 18:15 CEST

👾 DEV/TECH ARENA

LMDIS (Last Mile Delivery Intelligence System) is an AI-powered decision support tool for daily delivery fleet planning at Golilla. The system automatically analyses ~300 delivery addresses overnight, combining Google Street View imagery, OpenStreetMap road topology, and Claude AI vision to generate a structured operational briefing per order.
At its core, LMDIS resolves two simultaneous constraints: the minimum vehicle size imposed by cargo characteristics (weight, dimensions, product type — sourced via TMS API) and the maximum vehicle size permitted by road accessibility (carriageway width, turning radius, physical restrictions). The intersection determines the optimal vehicle assignment; when no intersection exists, the system identifies the nearest accessible point for the largest available vehicle and flags the conflict for manual resolution.
The road analysis engine follows actual OSM way geometry to sample points along both the delivery street and the access route from the nearest primary artery — eliminating the linear coordinate offset errors that make naive Street View analysis unreliable. A four-level heading cascade (OSM geometry → Google Roads snap-to-road → panorama inference → manual review flag) ensures no analysis proceeds with an incorrectly oriented image.

Speakers

Andrea Millimaggi

GoLilla

Phillip Rech

Product Leader, GoLilla

Phillip Rech I'm a Senior Digital & AI Manager at Golilla, where I lead digital transformation initiatives across logistics operations serving major retail clients including Leroy Merlin, Tecnomat, and Brico Center.
With a background spanning project management, enterprise architecture, and applied AI, I work at the intersection of operational complexity and emerging technology... Read More →

Tuesday May 26, 2026 17:45 - 18:15 CEST
👾 DEV/TECH ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Advanced

09:00 CEST

The 50 GB revelation: How a simple statistic redefined how we see data at Decathlon

Wednesday May 27, 2026 09:00 - 09:45 CEST

🤖 DATA/AI ARENA

For years, the industry's motto has been simple: collect and store as much data as possible, believing that immense value will emerge from this accumulation. This vision, driven by tech giants like **GAFAM** that collect petabytes of data daily, isn't always practical for every company. At the same time, single computing instances on the cloud have become exponentially more powerful, allowing us to handle large amounts of data on a single machine.

This new way of thinking is captured by the **Small Data Manifesto**, which states that more data doesn't always equal better results, modern hardware is often underused, and data topics should be developed locally first. In the data department of Decathlon, we discovered this is more than just a theory. More than 90% of the data tables in our datalake (Decathlon's centralized data repository) was less than 50 GiB. This finding inspired us to define a new approach to data transformation.

After setting the context and introducing the Small Data Manifesto, we will present **Light Computing**, which brings these principles to life. Instead of spinning up expensive clusters to transform data, we use powerful single-node tools and only scale up when the data truly requires it. The leading tools of this field are **Polars** and **DuckDB**. We'll cover two main aspects: what technically defines these tools and what they bring in terms of concrete cost and architectural benefits.

The session will conclude with a live demonstration of Polars, showing how we can easily compute 50 GB of data on a small computing instance on Databricks.

Speakers

Arnaud Vennin

Data Engineer, Decathlon

Raised on the eastern plateaus of Rouen in the northwest of France, I studied primarily in my hometown, where I completed my undergraduate degree in Environmental Sciences. I spent one semester abroad in Nijmegen, a city in the Netherlands. Afterward, I obtained a master's degree... Read More →

Wednesday May 27, 2026 09:00 - 09:45 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

09:15 CEST

MICROSOFT - The intelligence loop: turning real-time signals into autonomous retail decisions

Wednesday May 27, 2026 09:15 - 10:00 CEST

1 - MAIN STAGE

Retail organisations are awash in signals, transactions, inventory updates, promotions, and operational events—but turning those signals into timely, business‑relevant decisions remains challenging. The issue is not data volume, but how intelligence is structured, contextualised, and activated as conditions change in real time.
This session explores a systems‑level approach to retail intelligence by combining Microsoft Fabric Real‑Time Intelligence with Microsoft IQ across Operational, Analytical, and Agentic layers. Rather than focusing on individual use cases, the talk presents a layered intelligence model for designing adaptable retail decision platforms.
At the foundation of this model is Fabric IQ, built on graph technology through Fabric Graph, which provides the ontology and semantic layer for representing entities, relationships, and business context across the retail domain. This graph‑based foundation enables signals—captured through Real‑Time Intelligence as both a sensing and activation layer—to be interpreted consistently across systems and time.
On top of this semantic graph foundation, agent‑driven intelligence can reason over live business context and drive decisions and actions, creating a closed loop between signals, understanding, and execution. Together, real‑time streaming, graph‑based semantics, and layered intelligence form a reusable system pattern for building retail‑scale intelligence platforms without being tied to a single workflow or implementation.

Speakers

Devang Shah

Principal Program Manager, MICROSOFT

Wednesday May 27, 2026 09:15 - 10:00 CEST
1 - MAIN STAGE

Data / AI

09:45 CEST

From search bar to shopping list: How LLM turns DIY intent into guided commerce

Wednesday May 27, 2026 09:45 - 10:15 CEST

🤖 DATA/AI ARENA

RAGEdito was born from a simple observation: when a customer types "how to change a window" into the Leroy Merlin France search bar, they are not looking for a product — they are planning a project. Currently live in production on French traffic, RAGEdito generates real-time DIY tutorials with contextualised shopping lists, directly from the search bar.

This session delivers a hands-on retrospective from a three-person Product Owner / Data Analyst / Data Scientist team on building, steering, and continuously improving a RAG system in real production conditions.

Product angle: how to translate a user signal into a structured product vision — from defining meaningful business KPIs to aligning stakeholders across a multi-product Search & Publication domain. How RAGEdito fits into a coherent, scalable roadmap with group-wide deployment ambitions.

Analytical angle: how production query analysis identified a significant share of editorially-intentioned traffic going unaddressed, segmented customer needs by intent type (repair, installation, construction, inspiration), and converted these insights into concrete growth levers and roadmap decisions.

Technical angle: how user feedback feeds a custom evaluation pipeline, refines LLM generation, and drives architectural evolution toward a reusable API component scalable across the group.

Key takeaway: how to align product vision, analytical intelligence, and technical robustness to turn a RAG prototype into a product with measurable business impact.

Speakers

Zi Wang

Adeo Services

Software Engineer turned Data Scientist, banking -> fmcg -> retail.

Svetlana Vasiukova

Product Manager, ADEO Services

Ekaterina Rezanovich

Data Analyst, ADEO Services

Data Analyst. 8 years of experience in data analysis, ad-hoc analytics, and AI/LLM use cases. At Adeo Group since 2021.

Wednesday May 27, 2026 09:45 - 10:15 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

10:15 CEST

DBT LABS - The next phase of data platforms, picking the best compute engine for any task

Wednesday May 27, 2026 10:15 - 11:00 CEST

2 - GRAND PAVILION

Depending on a single vendor for all data needs creates significant bottlenecks: vendor lock-in, unoptimized compute engines for specific workloads, and slow developer feedback loops. The next phase of data architecture demands a shift toward decoupled layers and flexible compute.

In this session, we will explore the future of data transformation built around open table formats and multi-engine processing. We will discuss how open formats like Apache Iceberg can turn Data Mesh theories into reality, while addressing current catalog challenges and where the ecosystem is heading next.

Finally, we will focus on enhancing the developer experience through local compute. We will run a practical demonstration where a Databricks/BigQuery dbt project is executed locally on DuckDB using automated SQL dialect translation. The result: zero cloud compute costs during development and lightning-fast iteration cycles for your data engineering teams.

Speakers

Jeremy Cohen

DBT Labs

Benoit Perigaud

DBT Labs

Wednesday May 27, 2026 10:15 - 11:00 CEST
2 - GRAND PAVILION

Data / AI, Conference

Audience Beginner

10:15 CEST

KONG - Context is king: Why is context the key to unlocking the true value of agentic AI?

Wednesday May 27, 2026 10:15 - 11:00 CEST

1 - MAIN STAGE

Overview

As AI models converge on strong reasoning capabilities, the differentiator for effective
agentic systems is increasingly not the model itself, it's the context you provide it.
Context engineering — the discipline of delivering the right information, at the right
granularity, at the right time — is becoming a core competency for teams building agents
that work reliably in production. In this talk, we define the problem, share practical
principles for context architecture, and walk through how we've approached it with our
own products.

Outline
I. The Context Problem
- As agents take on longer, multi-step tasks with real-world consequences,
the quality of their context becomes the primary driver of output quality

- Most agent failures in practice aren't reasoning failures — they're context
failures: the model was missing information, working with outdated
information, or overwhelmed by irrelevant information

- Context engineering as a discipline: distinct from prompt engineering,
closer to systems design. It's about pipelines, infrastructure, and
information architecture — not just what goes in the system prompt.

II. What Good Context Looks Like
- Relevance is the measure that matters – The goal of context engineering
isn't to maximize information available to the model — it's to maximize
relevance. Everything else follows from this.

- You can't always predict relevance in advance – Agentic workflows are
dynamic. What's relevant at step five depends on what the agent discovered
at step three. Pre-curated context assumes you know the path before the
agent walks it.

- Explorability is how you achieve relevance at scale – Rather than
pre-identifying the right context, the more resilient approach is making your
knowledge base easy for the agent to traverse and explore on the fly. The
design question shifts from "build a better retriever" to "make your
knowledge navigable."

- This is where most teams are under-investing – The default
embed-index-retrieve pattern optimizes for static similarity. It doesn't
support an agent that needs to follow threads across systems, discover
adjacent context, or refine its understanding as a task unfolds.

- Measure whether agents are finding what they need – Without this, you
can't tell whether a bad output is a reasoning problem or a context problem.

III. Turning Enterprise Knowledge into Agent-Ready Context

- Enterprise environments are where context engineering is both most
diﬃcult and most impactful — the knowledge exists but it's scattered,
siloed, inconsistently structured, and constantly changing

- The untapped goldmine: docs, wikis, tickets, CRMs, codebases, Slack
threads — most enterprise knowledge is already there, just not accessible
to agents

- Strategies for ingesting, indexing, and surfacing enterprise knowledge at
inference time

- Handling permissions, access control, and data sensitivity in context
pipelines

- How we achieve this with Kong Context Mesh: a walkthrough of our
approach to transforming enterprise resources into agent-ready context

1/ Our architecture for ingesting, indexing, scoping, and serving
enterprise context at inference time

2/ Assumptions we made early on that turned out to be wrong, and
what we learned

3/ What measurably improved when we got context right, and where
we're still iterating

Speaker Bio

Christopher Tam is the Product GM for Kong’s Agentic AI Infrastructure offerings,
including AI Gateway, Context Mesh, and KAi. A veteran in the AI and infrastructure space,
Chris previously founded Substrates.ai, an agentic infrastructure startup, and held
product leadership roles at Google and Verily, where he applied cutting-edge AI
technologies to business applications. He was also a VP at Leap Motion, an early pioneer
in computer vision.

Speakers

Christopher Tam

Kong

Christopher Tam is the Product GM for Kong’s Agentic AI Infrastructure offerings,
including AI Gateway, Context Mesh, and KAi. A veteran in the AI and infrastructure space,
Chris previously founded Substrates.ai, an agentic infrastructure startup, and held
product leadership ro... Read More →

Wednesday May 27, 2026 10:15 - 11:00 CEST
1 - MAIN STAGE

Data / AI, Conference

Audience Intermediate

10:30 CEST

Smarter last-mile logistics: using operations research to optimize delivery districts and booking calendars in the Italian context

Wednesday May 27, 2026 10:30 - 11:00 CEST

🤖 DATA/AI ARENA

Last-mile delivery operations face increasing pressure to balance service quality with cost efficiency, given highly variable demand densities and heterogeneous territories. We present an optimization framework that jointly designs territorial delivery districts and customer booking calendars, treating delivery slot availability as decision variables. In remote or low-demand areas, the system restricts booking availability to a limited number of weekly delivery slots, enabling order consolidation and reducing inefficient travel. Conversely, high-demand districts are offered a larger set of booking slots, supporting more frequent delivery routes. **By shaping demand in time, the system aligns customer delivery choices with operationally efficient routing patterns**. The underlying mathematical formulation is based on a Mixed-Integer Linear Programming (MILP) model that optimizes booking calendar configurations while minimizing fleet size and traveled kilometers under company-specific operational constraints.
Deployed by Golilla, the Italian last-mile logistics operator within the Adeo ecosystem, the tool was tested over six months at a regional hub in Emilia-Romagna, replacing the previous experience-based scheduling approach. The result was a 15% increase in gross margin. It is now live in production and being rolled out across the full Italian territory.

Speakers

Dario Piloni

Data Scientist, Tecnomat Italy

Data scientist in Tecnomat Italy since 2023, formerly a scientific researcher in the field of Operations Research, Mathematical Optimization and Machine Learning.

Nicholas Mattia Marazzi

Lead data scientist, Tecnomat Italy

Nicholas Marazzi, lead data scientist, in Tecnomat since 2022, formerly a scientific researcher and data science consultant.

Wednesday May 27, 2026 10:30 - 11:00 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

10:30 CEST

Industrializing forecasting solutions: A modern architecture

Wednesday May 27, 2026 10:30 - 11:15 CEST

3 - LAB

At Decathlon, based on years of experience, we’ve completely revamped how we do forecasting by building a "forecasting pipeline factory." Thanks to this approach, we are now building and shipping pipelines in producion in less than two months. Our solutions are based on modern data manipulation tools (dbt), in-house python libraries for machine learning abstractions, and modern models such as Time Series Foundation Models (Chronos2).

This technical stack, combined with an “extreme programming” mindset, has greatly accelerated our developments and deployment velocity. It’s been a real cross-functional enabler to deliver actual business value across several units at Decathlon.

Speakers

Raphaël Nedellec

Staff Data, Decathlon

Staff Data at Decathlon, InStore domain.

Vianney Bruned

Decathlon

Vianney Bruned is a staff Data Scientist at Decathlon in the Value Chain domain.

Anthony Vromant

Decathlon

Anthony Vromant is an ML Engineer in the Decathlon Value Chain domain.

Wednesday May 27, 2026 10:30 - 11:15 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

11:00 CEST

GOOGLE - Gemini Workshop: Advanced generation use cases

Wednesday May 27, 2026 11:00 - 12:00 CEST

🏈 PRACTICE ZONE #3

Enhance your problem-solving toolkit by integrating multimodal and million-token context capabilities, and unlock an unprecedented range of new solutions.

In this interactive workshop, we'll address the following complex challenges using Gemini and Python notebooks:

Multimodal Video Transcription: Transcribe videos and identify speakers in a single prompt.
Knowledge Graph Generation: Extract entities and relationships from massive inputs (1M tokens) with a single request.
Image Bank Automation: Set up a pipeline for consistent image generation and bring your visual archives back to life.

No expertise, preparation, or installation is needed. Bring your laptop, a browser, and a pinch of curiosity!

Speakers

Laurent Picard

Developer Relations Engineer, Google

Wednesday May 27, 2026 11:00 - 12:00 CEST
🏈 PRACTICE ZONE #3 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

11:15 CEST

How to establish an agentic framework: Sovereignty and simplicity with "Small Agent"

Wednesday May 27, 2026 11:15 - 11:45 CEST

🤖 DATA/AI ARENA

We are hearing a lot about agentic AI right now. Every developer is testing it, looking to implement it, and some are even pushing it to production. However, the ecosystem is still highly uncertain; the surrounding technologies (ADK, LangGraph, CrewAI, etc.) are very young. Ultimately, picking one right now feels more like a gamble than an informed choice.

**So, what solution would allow us to both set a clear framework and ensure agentic sovereignty? Small Agent!**

Within the LMFR AI team, we have built a framework that empowers any developer to implement agents in just three lines of code. All the underlying complexity is managed by the framework itself. Thanks to A2A, we standardize the interfaces so that, in the end, whether you choose LangGraph, ADK, or another tool behind the scenes, it simply doesn't matter.

It features two operating modes:
* Low-Code: If you want minimal code to run a V0 agent, let the framework do the heavy lifting.
* Advanced: If you are experienced in orchestration, you have full control. For example, if you use LangGraph, you can manage the graph directly. Everything is defined as classes, leaving you completely free to use code overrides as your imagination dictates.

This framework will enable us to:
* Accelerate everyone's daily workflow.
* Centralize complexity within the AI team, allowing every developer to implement easily.
* Achieve true technological sovereignty.
* Guarantee agentic consistency across the company.

**Additional Notes / Context:**

And yes, this really exists! We are already starting to implement it in specific areas. We also have our sights set on deploying an operational agent directly onto Turbine, using the A2A protocol to standardize communications with the agents.

Speakers

Axel Tison

Data Scientist, Leroy Merlin France

DS @LMFR

Cyril Spanneut

Lead AI, Leroy Merlin France

Hi ! I'm Cyril,

Wednesday May 27, 2026 11:15 - 11:45 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

11:15 CEST

LINKUP - Designing a retrieval layer for AI agents: Lessons from production

Wednesday May 27, 2026 11:15 - 12:00 CEST

2 - GRAND PAVILION

Most AI systems today rely on Retrieval-Augmented Generation (RAG) or APIs to access external knowledge. While
effective in controlled environments, these approaches often fail in production when faced with real-world data: incomplete
coverage, stale information, and unreliable sources.
In this talk, we share practical lessons from building a web-scale retrieval layer designed specifically for AI agents. We’ll
explore the key challenges of working with web data, dynamic content, inconsistent structures, adversarial noise, and the
architectural decisions required to handle them.

Topics include retrieval strategies (search vs direct fetch), ranking signals (relevance vs trust), structuring unstructured data,
and the trade-offs between latency, precision, and coverage.
Through real-world failure cases and design patterns, attendees will gain a clearer understanding of how to build AI systems
that remain grounded, reliable, and useful beyond demos.

Speakers

Boris Toledano

COO, Linkup

Boris Toledano is the Co-founder and Chief Operating Officer of Linkup, a cutting-edge AI technology company dedicated to bridging the gap between Large Language Models and the real-time web.

Wednesday May 27, 2026 11:15 - 12:00 CEST
2 - GRAND PAVILION

Data / AI, Talk

Audience Intermediate

12:00 CEST

From crawled pages to strategic insights: AI-powered Competitive Insights at scale

Wednesday May 27, 2026 12:00 - 12:30 CEST

4 - MEDIUM STAGE

How do you know if your product assortment is complete, well-structured, and competitively relevant — across **millions of products**, updated weekly, without a team of analysts manually browsing competitor websites?

At Leroy Merlin (ADEO), we faced exactly this challenge. Category managers needed actionable competitive intelligence: a structured view of **assortment coverage** — which product types competitors offer, which attributes they highlight, and where gaps exist. That means extracting structured product attributes — dimensions, materials, capacities, product types — from raw competitor product pages and mapping them onto our own product taxonomy.

We built a fully automated, AI-powered pipeline running on Google Cloud (Vertex AI + BigQuery) that today monitors **31 retailers and marketplaces** — including ManoMano, Castorama, Amazon, CDiscount, Ikea, Bricorama, and even the Leroy Merlin marketplace itself — processing over **26 million products** weekly.

The system combines two complementary AI approaches:
- A **custom ML classification model** — the **PEM (Product Entity Matching) API**, an existing ADEO digital asset reused here for a new purpose — that assigns each competitor product to one of **2,873 product models** within ADEO's internal taxonomy. It successfully classifies **62.7% of all ingested products**, with confidence-score thresholding ensuring downstream data quality.
- **Gemini 2.0 Flash** for structured characteristic extraction from raw product text, using few-shot prompting, self-verification loops, and type-aware validation (closed value sets, numeric measurements with automatic unit conversion). This has produced over **39 million characteristic values** across **80 distinct attributes** on **11.5 million products**.

The results feed a **dbt aggregation pipeline** in BigQuery, stream to a **Vue.js frontend**, and give category managers a live, filterable view of how Leroy Merlin's catalog compares to the market — by product type, characteristic, and price quartile.

In this Tools in Action session, we'll walk through the full pipeline architecture, the AI design choices and their trade-offs (LLM output validation, reuse of internal ML assets, quality filtering), and close with a live demo of the app.

Speakers

Guillaume Caron

Data Scientist, Leroy Merlin France

Julien Momal

Product Manager, Leroy Merlin France

### Julien Momal — Product Manager, Leroy Merlin France

Julien has been a Product Manager at Leroy Merlin for over 4.5 years. As Leroy Merlin France Domain Lead for Offer & Purchasing topics, he bridges the gap between business impacts and innovative product delivery.

Wednesday May 27, 2026 12:00 - 12:30 CEST
4 - MEDIUM STAGE

Data / AI, Tools In Action

Audience Intermediate

12:00 CEST

Make your data pipelines more reliable: Alerting and automated calculation of SLIs through meaningful quality metrics

Wednesday May 27, 2026 12:00 - 12:30 CEST

🤖 DATA/AI ARENA

For the dataplatform teams, building high-performance pipelines is only a part of the equation: the real challenge is ensuring that this data is accurate, up to date and usable by business units and AI models on a daily basis. Data quality has thus become the true battle of trust between technical teams and end users.

However, ensuring this quality across an entire organisation is complex, therefore defining and enforcing global data quality rules across an entire company rarely works: only local teams understand the reality and specific nature of their data. Yet, allowing each team to manage its indicators in silos creates chaos in communication, forcing Product Managers (PMs) in particular to manually cobble together their Service Level Indicator (SLI) calculations and user alerts. An unsustainable situation, especially since data of impeccable quality is now an essential prerequisite for deploying artificial intelligence on a large scale.

In this talk, we will present our experience in creating a standardised data model. We will also see why quality measurement must remain the responsibility of the local product team, while standardising its reporting via a ‘Plug & Play’ architecture. We will the explain how to design this layer independently of the underlying technical solution in order to use these unified metrics to generate reliable SLIs. Finally, we will discuss the prospects offered by this formalism, which paves the way for more complex use cases such as feeding a cross-functional Data Cockpit or evaluating real Data Contracts.

The goal isn't to dictate a tool, but to create a universal standard of communication. We will detail this agnostic architecture that acts as the single source of truth for your pipelines. No more mental load for Product Managers: this overlay automatically transforms every test suite into business SLIs and proactive alerts, making your data reliability 100% autonomous.

Speakers

Thomas Fenot

Decathlon

As a Data Engineer at Decathlon for 4 years, Thomas builds and scales data products within the Data Platform Business Unit. Operating in the "Core Data Product Planet and Industry" team, his primary focus is delivering the complex data models required to calculate product carbon footprints... Read More →

Wednesday May 27, 2026 12:00 - 12:30 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

14:00 CEST

Define once, run anywhere: portable SQL pipelines for modern data platforms

Wednesday May 27, 2026 14:00 - 14:30 CEST

🤖 DATA/AI ARENA

SQL remains the most widely used language in data engineering.

Yet modern data platforms are increasingly fragmented: warehouses, query engines, lakehouses, streaming systems and AI pipelines often require different tools, different execution environments and different ways to define transformations.

What if SQL pipelines could be **portable across engines**, while also providing built-in guarantees for data quality and testing?

In this talk, I will introduce **RivetSQL**, an experimental project exploring a new approach to **declarative SQL pipelines with multi-engine execution**.

The goal is simple:

> Define your pipeline once, run it anywhere.

RivetSQL allows developers to describe data pipelines declaratively while enabling:

- execution across different SQL engines
- built-in data quality checks
- integrated testing of transformations
- reproducible and portable pipelines

In this session, we will explore:

- why SQL pipelines remain difficult to make portable
- how modern data stacks introduce new execution challenges
- the architecture behind RivetSQL’s multi-engine execution model
- how testing and data quality can be integrated directly into pipeline definitions
- lessons learned from building an early-stage language experiment

Finally, we will discuss how these ideas could impact **future data platforms and AI pipelines**, where portability, reproducibility and data quality become critical.

This talk is aimed at engineers building data platforms, analytics systems or AI pipelines who want to rethink how SQL pipelines should work in modern infrastructures.

Speakers

Massil Chabane

Decathlon

Staff Data Engineer at Decathlon, I have been working on data platforms and data engineering for the past 6 years, after starting my journey in tech more than 10 years ago.

At Decathlon, I contributed to building parts of the company’s data platform, including initiating the automated data ingestion tooling that powers Decathlon’s data catalog. I later led the data architecture team, helping shape how data platforms evolve at scale in a large retail... Read More →

DEVSUMMIT2026 RivetSQL pdf

Wednesday May 27, 2026 14:00 - 14:30 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Intermediate

14:00 CEST

GOOGLE - Creating a “Dungeon Master” with Postgres and MCP

Wednesday May 27, 2026 14:00 - 14:45 CEST

3 - LAB

Instead of building another boring chatbot, let's create a Dungeon Master for our next Dungeons & Dragons campaign! Using this practical and fun example, we will build an AI agent that runs entirely on PostgreSQL. We'll go beyond simple query generation to explore how to grant agents secure, contextual access to your database for complex, non-predictive tasks. You'll learn how to architect an MCP (Model Context Protocol) server to prevent rogue AIs from dropping your tables while still empowering them to act as creative partners.

Join this quest to save the realm of elephants and learn to forge the weapons you'll need for the coming run.

Speakers

Matt Cornillon

Specialist Customer Engineer, GOOGLE

With 10 years of experience as a database enthusiast, I now support developers, engineers, and DBAs at Google as they onboard and scale with cloud databases.

My work is deeply rooted in the open-source community, where I serve as Vice-President of PostgreSQL Fr and Significant PostgreSQL Contributor. I remain passionate about the evolution of database engines and helping teams leverage the power of Postgres in the cloud

... Read More →

Wednesday May 27, 2026 14:00 - 14:45 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

14:00 CEST

GOOGLE - Retail in Agentic era

Wednesday May 27, 2026 14:00 - 14:45 CEST

1 - MAIN STAGE

Speakers

Mark Steel

Director, Global Retail & Commerce Industry Strategy, Google

I am an experienced Retail & Consumer industry leader with a proven track record of leveraging technology, AI & digital transformation to deliver significant revenue growth I have strong experience in the retail and wider consumer sectors across a diverse mix of businesses at different... Read More →

Wednesday May 27, 2026 14:00 - 14:45 CEST
1 - MAIN STAGE

Data / AI, Conference

14:00 CEST

GOOGLE - Gemini Workshop: Advanced generation use cases

Wednesday May 27, 2026 14:00 - 15:00 CEST

🏀 PRACTICE ZONE #1

Multimodal Video Transcription: Transcribe videos and identify speakers in a single prompt.
Knowledge Graph Generation: Extract entities and relationships from massive inputs (1M tokens) with a single request.
Image Bank Automation: Set up a pipeline for consistent image generation and bring your visual archives back to life.

No expertise, preparation, or installation is needed. Bring your laptop, a browser, and a pinch of curiosity!

Speakers

Laurent Picard

Developer Relations Engineer, Google

Wednesday May 27, 2026 14:00 - 15:00 CEST
🏀 PRACTICE ZONE #1 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

14:00 CEST

Hands-On DBT

Wednesday May 27, 2026 14:00 - 15:30 CEST

🏈 PRACTICE ZONE #3

Scaling dbt projects requires moving beyond basic assertions. In this 90-minute hands-on workshop, Benoit from dbt Labs will guide you through industrializing your development workflow. We will dive deep into advanced testing strategies and build robust CI/CD pipelines using state-based execution. Learn how to automate confidence in your code and ensure every pull request is truly production-ready. Laptop required.

Speakers

Benoit Perigaud

DBT Labs

Wednesday May 27, 2026 14:00 - 15:30 CEST
🏈 PRACTICE ZONE #3 135 Rue Sadi Carnot, Ronchin, France

Data / AI

14:45 CEST

Cleaning the data lake without drowning: A PM & DataOps story

Wednesday May 27, 2026 14:45 - 15:15 CEST

🤖 DATA/AI ARENA

Storage is often treated as an infinite, invisible commodity, until the bill explodes or the Data Lake turns into a swamp. At Decathlon's scale, managing storage isn't just an infra topic; it's a **Product** that requires strategy, value measurement, and a serious cleanup.

In this session, **Lucie** (Senior PM) and **Soufian** (DataOps Engineer) share their unfiltered journey of reclaiming control over their storage zones. From the first discovery interviews to the automated deletion of Petabytes of data, they’ll reveal how they teamed up to turn a growth crisis into a sustainable efficiency model.

In this session, they will provide a deep dive into the practical synergy between:
- **Product Management for Infra**: How to apply PM frameworks to technical topics like storage costs.
- **The DataOps Engine**: A look under the hood of their automated cleanup stack (S3 Lifecycle, Delta Vacuum) and how they mitigated risks.
- **Cultural Shift**: How they moved from "keep everything" to "smart archiving" by design, proving that deleting is also delivering.

Whether you're a PM or a Dev, you'll leave with concrete tips on how to apply Product methodology to technical debt and ensure your Data Lake stays healthy and ready for the AI era.

Speakers

Soufian Salhi

Decathlon

As Cloud Infrastructure Engineer, Soufian has been contributing to the technical storage layers since 2022 for Decathlon’s Data Factory. For the past year, he has focused on the Data Lake's efficiency, developing the automated frameworks and cloud environments needed to manage data... Read More →

Lucie Bailly

Decathlon

Senior Product Manager within Decathlon’s Data Factory, Lucie has been tackling data challenges since 2012. Over her 14-year career, she has explored the various dimensions of data, from BI and Data Engineering to Data Governance, at leading scale-ups such as Criteo and Doctolib... Read More →

Wednesday May 27, 2026 14:45 - 15:15 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

15:00 CEST

Decoding customer value with Causal ML

Wednesday May 27, 2026 15:00 - 15:30 CEST

3 - LAB

Most retail models fail by confusing correlation with causality. High-value customers might create more projects, but does creating a project actually increase their value? This presentation showcases the Adeo XP Award-winning "Drivers of LTV" project, which isolated the true incremental impact of driver—from life events to digital interactions—to build a high-precision business simulator.

Speakers

Jiqiong Qiu

Data Scientist, ADEO Services

Jiqiong Qiu is an experienced data science professional who currently serves as a Lead Data Scientist at a customer commerce platform. With a strong background in leveraging data-driven insights to enhance customer experiences and drive business growth, Jiqiong specializes in developing... Read More →

Maxence Debert

Product Owner, ADEO Services / Hubvisory

Product Owner of the Client Scoring Platform (ADEO-CCDP-CKA), where I’ve spent the past two years building this AI/ML product that turns customer data into actionable scores and segments for client knowledge and marketing activation.As a consultant at Hubvisory, I enjoy shaping... Read More →

Frederic Verhasselt

ADEO Services

Wednesday May 27, 2026 15:00 - 15:30 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Advanced

15:00 CEST

ZALANDO - Powering the future of fashion through AI

Wednesday May 27, 2026 15:00 - 15:45 CEST

1 - MAIN STAGE

This keynote will explore how Zalando positions itself as a tech company, with AI at the core of its strategy.
It will highlight how AI is leveraged to elevate customer experience, driving differentiation through personalization, inspiration, and seamless journeys.

Speakers

Laura Toledano Khelif

General Manager, Western Europe, Zalando

https://www.linkedin.com/in/laura-toledano-khelif-19802a2b/

Wednesday May 27, 2026 15:00 - 15:45 CEST
1 - MAIN STAGE

Data / AI, Conference

Audience Advanced

15:30 CEST

From traditional search to knowledge graph-powered engine: The Leroy Merlin Brazil quotation generator experience

Wednesday May 27, 2026 15:30 - 16:00 CEST

🤖 DATA/AI ARENA

The Quotation Generator is an innovative solution designed to fully automate the creation of customer quotes, transforming a traditionally manual, time-consuming process into a seamless digital experience. At the heart of this initiative lies a critical challenge: accurately interpreting free-text inputs and recommending the right combination of products to build a complete and reliable quote.

In the second half of 2025, with the strategic support of Adeo and in partnership with TheOdo, we explored different approaches to address this challenge. In particular, we experimented with several search and matching techniques, including knowledge graph–based methods, to evaluate their potential for improving product recommendation and quote generation.

In this presentation, we will describe the experiments we conducted, the evaluation frameworks we developed to measure performance, and the results obtained across the different approaches. We will also share the key insights and lessons learned from this exploration, including cases where traditional methods proved more effective than more complex alternatives and which experiments performed better in a production environment.

Speakers

Lucas Eduardo De Cesar Morais

Data Scientist, Leroy Merlin Brazil

Lucas is a **Data Scientist at Leroy Merlin Brazil** and a **Pure Mathematics undergraduate at the University of São Paulo (USP)**. He has worked on several AI initiatives within the company’s digital ecosystem.

He started three years ago, as an intern, working on the **Sales C... Read More →

Leandro Marcelo Domingues Galvao

Data & AI Manager, Leroy Merlin Brazil

Since 2022, Leandro Galvão has been part of Leroy Merlin Brazil, where he currently serves as Data & AI Manager. In this role, he leads key initiatives across multiple areas, including Offer, Supply, and Marketing, leveraging his expertise in data strategy, team leadership, and process... Read More →

Wednesday May 27, 2026 15:30 - 16:00 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

15:45 CEST

From POC to production: Our product RAG agent journey

Wednesday May 27, 2026 15:45 - 16:15 CEST

3 - LAB

We are going to share the story behind the development and deployment of our Product RAG Agent, a conversational assistant integrated into our Product Detail Pages (PDP) to answer customer inquiries.

The project has been a clear success, driving a 10% increase in average basket value.

Driven by a mindset of continuous improvement, our development cycle was punctuated by three successful A/B tests between July and December. At the core of our architecture lies the vectorization of product manuals and packaging data. However, moving to production revealed unexpected challenges, specifically regarding latency issues and skyrocketing costs.

We will detail how we rapidly tackled these SLO failures and pricing concerns by enhancing our observability.

This session is a comprehensive feedback loop: come discover how we are scaling our architecture to meet tomorrow's challenges.

Speakers

Estelle Godard

Full-stack Developer, ADEO Services

Hey ! I'm Estelle, I am a Full-stack Developer at ADEO for over two years, I started this journey with a one-year apprenticeship before transitioning into a permanent role. I am currently a member of the Customer Decision team within the Search & Publication domain, working on the... Read More →

Baptiste Lecocq

Ext - ADEO Services

Wednesday May 27, 2026 15:45 - 16:15 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

16:15 CEST

Would you ship code without tests? Why do your AI agents' skills need evals

Wednesday May 27, 2026 16:15 - 16:45 CEST

🤖 DATA/AI ARENA

Everyone is writing skills. Almost nobody is testing them and most of them are AI-generated.
Skills might get "vibe-checked" with a handful of manual runs, then shipped.

## You wouldn't ship code without tests, but why ship skills without evals?

As we transition from simple chat interfaces to autonomous AI Agents, equipping LLMs with tools (APIs, functions) has become the new standard.

This talk tackles the critical missing piece in Augmented Development: Skill Engineering and its evaluation.
We will move past the "vibe check" and dive into the methodologies required to build robust, measurable agents based on industry best practices.

## What you will learn in this 30-minute session:

- LLM-Friendly design: How to write semantic schemas and tool descriptions that models actually understand, reducing baseline errors.
- TDD for AI agents: How to define success criteria and build automated tests for non-deterministic systems.
- The evals playbook: Measuring what matters by focusing on routing accuracy (did it pick the right tool?) and argument accuracy (are the parameters valid?), including how to leverage "LLM-as-a-Judge".
- Continuous refinement: Using failed evals and production telemetry to iteratively improve your skill prompts without touching the underlying business logic.

## Stop guessing if your agents work.

Join this talk to learn how to test, measure, and refine your AI skills with the same rigor as traditional software engineering.

Speakers

Thomas D'hulst

Mobile software engineer - CMA, Ippon Technologies

Wednesday May 27, 2026 16:15 - 16:45 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

16:30 CEST

Agentic AI: Beyond the hype, into the business

Wednesday May 27, 2026 16:30 - 17:00 CEST

3 - LAB

In the world of retail, the "Daily Brief" for a Store Manager is a high-stakes moment of truth. Whether at Adeo or Decathlon, the challenge is identical: how to distill a mountain of data into 15 minutes of operational action. To solve this, both companies launched ambitious AI initiatives—yet our journeys, technical choices, and "Aha!" moments often differed.
In this first-of-its-kind co-presentation, the Adeo and Decathlon AI teams join forces to share their parallel quests in building intelligent assistants. We will move beyond the hype of "Agentic AI" to discuss pragmatism. While Agentic workflows offer incredible power, they come with high costs and complexity. Is a "Ferrari" always the right choice for every retail feature, or can we deliver massive value with more economical, "lean" AI models?
What we will share:
Two Paths, One Goal: A comparison of our architectures—where we aligned and where our specific retail contexts (DIY vs. Sports) forced us to diverge.
The ROI of Intelligence: How we balance the high cost of sophisticated Agents with the need for scalable, low-latency business value.
The Shared "Grit": Common failures in data reliability, interfacing with legacy systems, and the reality of deploying AI in a fast-paced store environment.
This session isn’t about perfect slides; it’s a transparent look at how two of the world's leading retailers are navigating the shift from "Data-Heavy" to "AI-Empowered" operations.

Speakers

Sébastien Staes

Decathlon

With my teams, we define and drive the data strategy and delivery for the domains Finance and HR.
Be focus on the user needs to provide the best data solutions to reply to their needs!

Pauline Vilpini

Data Lead Expert, Leroy Merlin France

Victor Marchal

Leroy Merlin France

Victor MARCHAL
ML Engineer @LMFR

Wednesday May 27, 2026 16:30 - 17:00 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Intermediate

16:45 CEST

DOCTOLIB - When technology reinvents healthcare at scale with AI

Wednesday May 27, 2026 16:45 - 17:30 CEST

1 - MAIN STAGE

The demand for care is accelerating - driven by population growth, longer life expectancy, and the rise of chronic and mental health conditions. Healthcare systems are under increasing pressure.
At Doctolib, we believe technology - powered by AI - can fundamentally transform how care is delivered, improving both the daily lives of healthcare professionals and the health of millions of people. Today, Doctolib connects 80 million patients and 900,000 healthcare professionals across Europe, with AI becoming a core layer of the entire experience.

In this keynote, Julie Touyarot, VP Growth, shares a behind-the-scenes look at Doctolib’s growth strategy: how AI and an integrated product experience are reshaping practitioners’ daily workflows - and how the patient app is evolving into a trusted health companion.

A candid perspective on the strategic choices, the challenges, and what it really takes to scale a healthcare platform with AI at its core.

Speakers

Julie Touyarot

VP Growth & Marketing, Doctolib

Julie Touyarot has 15 years of marketing experience in tech. After building her career at Google in the US and France, she joined Uber in 2015. As Marketing Director EMEA for Uber and Uber Eats, she led five key markets, driving double-digit growth and positioning the brands as category... Read More →

Wednesday May 27, 2026 16:45 - 17:30 CEST
1 - MAIN STAGE

Data / AI, Keynote

17:00 CEST

Your language is the new code: How CPQ turns your words into a working configurator

Wednesday May 27, 2026 17:00 - 17:30 CEST

🤖 DATA/AI ARENA

Building a product configurator at Adeo has historically been a months-long endeavor: modeling compatibility rules, translating business constraints into code, endless test cycles… And evolving an existing rule? Days of work for the tech team.
With our CPQ (Configure, Price, Quote) project, **winner of the Adeo XP Awards 2026 in the Disruptive Innovation category**, we flipped this paradigm entirely. The only skill you need to use CPQ is the ability to write in your own language. A business expert describes a rule in plain words *"if the customer picks a floorboard wider than 15cm, only show compatible underlays"* and Gemini automatically generates the execution logic, along with a plain-language explanation to review before going live.
What used to take weeks now takes minutes. In this talk, we'll share how we built this AI pipeline (Spring AI + Vertex AI), the challenges we faced, and the lessons learned from a project where AI isn't a feature — it's the foundation.

Speakers

Theo Hamon

UX Designer, ADEO Services

Pierre Gerard

Lead Engineer, ADEO Services

Pierre GERARD, Lead Engineer at Adeo, where I have been building and scaling products for the past several years. I've been leading the CPQ project since day one — from its initial architecture to the AI-powered features that earned the team the XP Awards 2026 in the Disruptive... Read More →

Wednesday May 27, 2026 17:00 - 17:30 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Beginner

17:15 CEST

Get Your Business insights in minutes, not days (by talking to Your data with AI)

Wednesday May 27, 2026 17:15 - 17:45 CEST

3 - LAB

The next step in our data journey at Adeo is about making our insights as dynamic as our daily decisions.

Today, the time between asking a business question and receiving an answer can still span days—creating a "latency" that slows our momentum. While our dashboards provide the essential, trusted foundation for monitoring our long-term performance, we often encounter specific, tactical questions in our daily operations that require a more immediate touch.

In this session, the Supply Chain Digital Platform (SCDP) team and the Central Dataviz Global Tech and Data Platform (GTDP) team join forces to showcase how Conversational Analytics in Looker, powered by AI, enriches our existing analytics ecosystem by adding a new, intuitive layer of interaction.

The heart of this agility lies in Looker Semantic Layer. By following the principles of the Aggregated Data Products strategy, we ensure that this conversational interface is built on a consistent, and actionable foundation. It’s about giving our data a voice—one that speaks the same language as our business, with the reliability that Adeo standards demand.
What we will explore together:

Fluid Interaction: How natural language can surface operational insights in seconds, seamlessly bridging the gap between a thought and an action.
The Power of Aggregated Data Products: Why our unified data framework is the secret to a conversational experience that is both trusted and hallucination-free.
Empowering the Organization: How shifting toward "data dialogues" allows every team to be more autonomous, while enabling our Data experts to focus on deeper, high-impact strategic initiatives.
Beyond the Dashboard (But Not Without It): Redefining Self-BI through Conversational Analytics without losing our common ground. We will discuss the right place for the right usage, showing how conversational interactions work in parallel with official reports that remain essential for monitoring shared KPIs."

Join us to see how we are moving towards a data culture where our trusted insights are always just a conversation away and discover if most of your dashboards will disappear.

Speakers

Loic Miguet

Lead Data Analyst, Adeo Services

Hi!
I'm Loïc, Lead Data Analyst in the Supply Chain Digital Platform at Adeo!
I've been working at Adeo for 4 years and before that at Les Mousquetaires for 1,5 year.

Nathan Michel

Business Intelligence Specialist, ADEO Services

I am Nathan, a Looker Expert Consultant and Business Intelligence Specialist at ADEO. Since January 2025, I’ve been part of the GTDP, bringing over four years of experience in Looker and various Data & Analytics domains. I have a strong interest in exploring all facets of data—from... Read More →

Wednesday May 27, 2026 17:15 - 17:45 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI

17:45 CEST

Empirical evidence for alignment faking in a small LLM and prompt-based mitigation techniques

Wednesday May 27, 2026 17:45 - 18:15 CEST

👾 DEV/TECH ARENA

I work as AI Governance Lead at Decathlon with a backgrouns in responsible AI and AI safety research engineering. In this talk, I'll present a paper that I've published with NeurIPS and AAAI "Empirical Evidence for Alignment Faking in a Small LLM and Prompt-Based Mitigation Techniques". I have presented a similar talk at the AAAI Fall Symposium Series last year. Given the audience at this summit, I can also spend some time diving into the importance of AI safety in multinational organisations and how we can go beyond policy, to include technical AI safety measures.

Speakers

Jeanice Koorndijk

Senior AI Governance Lead, Decathlon

Senior AI Governance Lead @ Decathlon. AI Safety Research engineer & Machine learning and social scientist.

Wednesday May 27, 2026 17:45 - 18:15 CEST
👾 DEV/TECH ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Beginner

17:45 CEST

From spaghetti scripts to robust pipelines: Streamlining marketing data ingestion at Decathlon with DLT

Wednesday May 27, 2026 17:45 - 18:15 CEST

🤖 DATA/AI ARENA

Data engineers often face a dilemma: use heavy, low-code ETL platforms or maintain a mountain of custom Python scripts for REST APIs and SFTP servers. At Decathlon, we chose a third way.

In this "Tool in Action" session, we will demonstrate how we leveraged dlt to simplify the ingestion of Marketing, SEO, and SEA data from fragmented sources (S3, GCS, SFTP, and various REST APIs) into our Datalake. We will walk through a live-coding-style demo showing how to:
- Turn a simple Python generator into a production-ready pipeline in minutes.
- Handle automated schema evolution and data typing without manual DDL.
- Implement robust monitoring and observability to ensure data quality at scale.

If you are looking to replace "manual" boilerplate code with a Pythonic, "ETL-as-code" approach, this talk is for you.

Speakers

Pierre Monnet

Decathlon

Wednesday May 27, 2026 17:45 - 18:15 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Intermediate

17:45 CEST

Ghost stockout detection: When sales interruptions should raise eyebrows

Wednesday May 27, 2026 17:45 - 18:15 CEST

4 - MEDIUM STAGE

We call ghost stockout the situation in which the stock is marked as available on digital stock management inventories, but not physically available in the store.

We present an unsupervised algorithm for detecting ghost stockouts using only daily sales data. We model each SKU's daily sales outcome as a time dependent Bernoulli process with a parameter representing the probability of selling one product on a given day, denoted as p₀(t). A ghost stockout manifests as an anomalous zero sales streak, whose probability

P(N,t) = ∏ from i=1 to N of (1 - p₀(tᵢ))

serves as a detection signal. In a ghost free environment this signal is uniformly distributed across SKUs; in the presence of ghost stockouts, the empirical distribution exhibits a relative abundance of low P(N,t) values far exceeding the theoretical uniform baseline, a signature that is both theoretically predicted and experimentally confirmed.

We estimate p₀(t) via an adaptive moving average procedure that self calibrates its window size to maintain a bounded relative estimation error across the full range of SKU sale rates and seasonal profiles. Crucially, the entire pipeline, including p₀(t) estimation, error bounding and signal computation, was implemented in pure SQL on BigQuery, eliminating the need for dedicated ML infrastructure and dramatically reducing deployment cost.

Validated across two retail stores over one month (3,220 physical inventory checks), the method achieved an overall true positive rate of 15.5%, rising to 17.8% on self service departments. In the top performing department store pairs the TPR reached 35%, demonstrating that when applied to structurally well defined availability contexts, this fully unsupervised algorithm delivers detection performance competitive with supervised approaches.

Speakers

Pierre Leder

ADEO Services

Cesare Paulin

Data Scientist, Tecnomat Italy

Cesare Paulin is a data scientist and physics-informed AI researcher based in Milan, Italy. He specializes in integrating machine learning, signal processing, and mathematical modeling to tackle complex industrial and scientific problems, with applications in time-series forecasting... Read More →

Wednesday May 27, 2026 17:45 - 18:15 CEST
4 - MEDIUM STAGE

Data / AI, Talk

Audience Advanced

09:00 CEST

On-Call at 3 AM: How does our AI resolve incidents before you do?

Thursday May 28, 2026 09:00 - 09:30 CEST

🤖 DATA/AI ARENA

What if your best SRE was an AI capable of analyzing millions of signals in seconds?

In this session, I’ll show you how we built a "Home-Made" Incident Investigator to automate root cause analysis from the ground up. By leveraging the Model Context Protocol (MCP) to unify Datadog telemetry, GitHub source code, and ServiceNow tickets, our autonomous agent doesn’t just report issues—it investigates them.

We will dive into how the agent iteratively forms hypotheses, "peeks" into raw logs, and suggests code fixes the moment an alert fires. No more cognitive overload or endless manual digging during high-pressure outages. Come and learn how we are drastically slashing our MTTR (Mean Time To Recovery) by adopting an "Agent-First" approach to production operations.

Speakers

Quentin Delignon

Lead Software Engineer, ADEO Services

Quentin Deligon (Adeo Services)
Lead Software Engineer in Observability team in Adeo. Innersource and OpenSource enthusiast.

Sebastien Becker

Project Leader, ADEO Services

I am currently the Observability Lead Engineer at ADEO, but I spent several years 'in the trenches' as an SRE at CCDP. My daily reality involved cross-functional management of critical incidents (P1/P2) in the dead of night.

It was this hands-on experience, combined with the operational fatigue of late-night interventions, that drove me to design and develop Incident Investigator AI. As the lead developer of this tool, my goal is simple: to translate years of manual investigation expertise into an autonomous... Read More →

Thursday May 28, 2026 09:00 - 09:30 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Beginner

09:45 CEST

Breaking the bottleneck: Boosting AI inference for production

Thursday May 28, 2026 09:45 - 10:15 CEST

🤖 DATA/AI ARENA

Scaling complex AI deployments demands overcoming severe latency and hardware bottlenecks, especially for resource-intensive GenAI and Deep Learning models. This session dives into the OpenVINO toolkit to demonstrate how you can seamlessly bridge the gap between trained models and high-performance production environments. You will learn practical, code-level strategies to convert, optimize, and accelerate inference for both generative and conventional AI models across today's most popular frameworks.

Speakers

Ahmed Sanaa

Data Scientist, ADEO Services

I am a Senior Data Scientist at Adeo Productions (AI Factory squad) and a graduate of Télécom Paris. With over 6 years of experience, I specialize in the end-to-end delivery of AI solutions—from strategic scoping to full-scale deployment. My expertise spans Machine Learning, Deep... Read More →

Adrien Legrand

SFEIR

ML Engineer chez Sfeir depuis début 2024, je suis titulaire d'un doctorat en ML appliqué.
Passionné par l’IA au sens large, j’ai eu l’occasion de travailler aussi bien sur une variété de POCs, allant du clustering d’images à la détection d’anomalies, que sur l’industrialisation de pipelines. J’aime lever des questions, des haltères, le coude... Read More →

Thursday May 28, 2026 09:45 - 10:15 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Intermediate

10:30 CEST

Fine grained data provenance with Apache Spark

Thursday May 28, 2026 10:30 - 11:00 CEST

🤖 DATA/AI ARENA

Decathlon's Data Lake is organized into progressive layers that transform data through increasing levels of complexity to power reporting, visualization (e.g., on interactive dashboards), and eventually advanced Machine Learning & AI (e.g., product recommendation, demand forecasting, dynamic pricing, …). To achieve this, we build and maintain complex, distributed pipelines written in SQL, and we leverage Apache Spark’s engine to handle Big Data processing at scale on multi-node clusters. However, complexity comes at a cost: as we stack more and more data transformations, manually tracing the exact origin of a specific data item becomes increasingly difficult and unmanageable, creating a critical need for an automated solution. We have recently recruited a Data Engineer intern and partnered with academic experts from ENS - PSL and Université Grenoble Alpes to prototype a (Fine-Grained) Data Provenance tool compatible with Apache Spark.
The ability to track the provenance/lineage of granular data portions is critical for:
- Trust & Reliability: guaranteeing the accuracy of results for data consumers.
- Root Cause Analysis: diagnosing anomalies (e.g., aberrant turnover figures) to pinpoint the exact source of a problem.
- Impact Analysis: predicting how data updates will propagate through our versioned datasets.
- GDPR compliance: ensuring that sensitive data (PII) does not unintentionally "leak" into refined datasets.
- Testing: extracting representative subsets of data for lightweight integration tests and prototyping.
In this talk, we will present few concepts of data provenance and present where we currently stand and what we plan to build in the future.

Speakers

Ronan Fruit

Decathlon

Thursday May 28, 2026 10:30 - 11:00 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

11:15 CEST

Transformer-Based Reranking: Bringing semantic intelligence to E-commerce

Thursday May 28, 2026 11:15 - 11:45 CEST

🤖 DATA/AI ARENA

Offering the right product in the right place: the Holy Grail of e-commerce. Yet, in a catalog of millions of items, a bestseller can become "noise" if it appears out of context. While popularity has long dominated ranking systems, it has now hit a glass ceiling. How do we move from purely statistical sorting to more intelligent navigation?

Within the AAAI team at ADEO, we reached a radical conclusion: behavioral signals (clicks, sales) modeled by classic ML are no longer enough. They are semantically blind. The result? A category can become flooded with accessories or unrelated products simply because they are cheaper and highly clicked, ultimately breaking the browsing experience.

In this session, we will take you behind the scenes of our hybrid ranking architecture. We will detail how we integrated the power of Transformers into the heart of our engine to re-inject context where algorithms were once "deaf" to the meaning of words.

On the Agenda:
The End of "Click Dictatorship": Why behavioral ML alone creates harmful popularity bias.

Semantic AI in Action: How we use Transformers to map user navigation intent to catalog reality.

Hybrid Architecture: Making the precision of a Cross-Encoder coexist with the scalability of a ranking model (like XGBoost).

Results & Lessons Learned: Real gains measured in ranking quality, limitations encountered, and takeaways from deploying on a massive scale.

Techniques covered: Python, ZenML, Transformer Models, Statistics, NDCG, ML

Target Audience: Data Scientists, ML Engineers, Product Owners

Level: Intermediate

Speakers

Alex Lenfant

Data Scientist, Ext - ADEO Services

Alex LENFANT – Data Scientist
A former flight test engineer in the aerospace industry turned AI enthusiast, Alex traded Airbus cockpits for ranking algorithms. While he no longer lands planes, he now ensures that your browsing in the "Drills" category doesn't go into a tails... Read More →

Thursday May 28, 2026 11:15 - 11:45 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Talk

Audience Intermediate

11:15 CEST

Artificial Intelligence at Pellenc ST: from experimentation to industrial reality

Thursday May 28, 2026 11:15 - 12:00 CEST

3 - LAB

At Pellenc ST, artificial intelligence is deployed directly into industrial systems operating in real-world waste sorting environments. This talk shares practical feedback on transforming AI prototypes into reliable, production-ready solutions with measurable ROI. We will present concrete use cases combining computer vision and spectrometry, and discuss the challenges of working with complex, noisy data collected under harsh industrial conditions. The session will cover key topics such as model robustness, edge deployment constraints, and system integration within high-performance machines. A strong focus will be placed on operational excellence: monitoring model drift, ensuring observability, and maintaining performance over time. We will also share insights on cost-performance trade-offs and technology choices in an industrial context. Finally, we will highlight how AI can contribute to sustainability by improving sorting efficiency and material recovery — demonstrating that AI can be both economically and environmentally impactful.

Speakers

Kevin Alazet

AI Development and Industrialization program manager, Pellenc ST

AI Development & Industrialization Program Manager at Pellenc ST, I am responsible for bringing artificial intelligence from prototype to production in industrial waste sorting systems. I work on deploying computer vision and spectrometry-based models directly into real-time machines... Read More →

Thursday May 28, 2026 11:15 - 12:00 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

11:15 CEST

PHOTOROOM - State of AI revolution in imaging: How to not become Polaroid or Kodak

Thursday May 28, 2026 11:15 - 12:00 CEST

1 - MAIN STAGE

State of the AI Revolution in Imaging: Don't Become Polaroid or Kodak

in 1975, Steve Sasson at Kodak invented the digital camera. Polaroid pioneered instant photography. Both went bankrupt, not because they missed the technology, but because they refused to redesign their operating model around it. We are living through the same inflection point in imaging right now, and most enterprises are making the same mistake.
In this talk, Jeff Strauss draws on Photoroom's work with global marketplaces and brands: Mercari, Valuence Japan, Decathlon, British Red Cross, Door Dash, Wolt, to show why AI imaging is not only a creative tool, it is operational infrastructure. He'll explain why pilots plateau, why "better models" aren't the answer, and what the iterative flywheel looks like when it actually compounds.

A central question runs through the talk: how do you measure "good" once humans are no longer reviewing every image? Jeff introduces Photoroom's three pillars of visual excellence — Fidelity, Realism, and Photographic Integrity — the framework that turns subjective creative judgment into a system any team can scale, audit, and improve against. Fidelity ensures the image is true to the product. Realism ensures it's believable to the buyer. Photographic Integrity ensures it holds up to the standards of professional craft. Together, they give leaders a defensible answer to the question every AI imaging program eventually faces: "How do we know this is working?"

Attendees will leave with a clear playbook: how to translate creative guidelines into enforceable presets without producing visual sameness, the metrics to track from day one, and the governance moves that protect buyer trust as you scale. The teams that treat AI imaging as a creative experiment will keep running pilots. The teams that treat it as infrastructure, measured against a clear definition of visual excellence will define the next decade of e-commerce, and leave the Kodaks and Polaroids of imaging behind.

Speakers

Jeff Strauss

Head of Imaging, PHOTOROOM.com

Over 35 years of experience in the photographic industry with 25 years building and managing large corporate photography teams around the globe. I now am a subject matter expert at Photoroom talking about the state of AI image editing and what is happening in our market as AI is... Read More →

Thursday May 28, 2026 11:15 - 12:00 CEST
1 - MAIN STAGE

Data / AI

12:00 CEST

From KPI Chaos to GenAI-readiness: How we structured Analytics at Decathlon and centralized KPI definition "as-code" in a federated Semantic & metrics layer

Thursday May 28, 2026 12:00 - 12:30 CEST

🤖 DATA/AI ARENA

Decathlon's decentralized analytics organization faced **KPI Chaos**, which was a bottleneck for next-gen BI and **AI-readiness**. This talk explores how establishing an **Analytics Center of Excellence (CoE)** enabled the delivery of a **federated Semantic layer**.

The Analytics CoE is driven by **Staff Engineers for Analytics**. Using **transversal leadership** and a **communities-based model**, we were able to source bottom-up bandwidth for transversal projects without dedicated budgets or direct hierarchical authority.

This empowered the delivery of `insight-all-metrics`, a **federated Semantic layer** built on **Databricks Metric Views** and **dbt**. Domains centrally expose their shared KPIs via simple YAML files. Then, any data practitioner can easily compute those KPIs. Leveraging **materialization features** reduced compute costs and enabled the instantaneous querying required for real-time GenAI applications.

By gathering together the **Metric Layer** (SQL logic to compute KPIs) and the **Semantic Layer** (natural language context) within a single repository, we bridged the gap between human questions and database queries. This single governed source of truth now powers both **Tableau Pulse** and **Databricks Genie**, marking our successful transition to **GenAI for BI**.

I will live demo some features.

Speakers

Sébastien Staes

Decathlon

With my teams, we define and drive the data strategy and delivery for the domains Finance and HR.
Be focus on the user needs to provide the best data solutions to reply to their needs!

Hugo Palmer

Decathlon

Thursday May 28, 2026 12:00 - 12:30 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

12:00 CEST

Road to sparse encoders: Bridging product catalogs and knowledge graphs

Thursday May 28, 2026 12:00 - 12:30 CEST

4 - MEDIUM STAGE

In the retail industry, the accuracy of product to concept mapping is the backbone of customer experience. At Leroy Merlin France, this represents a massive operational challenge: over 5.7M product-concept pairs (out of 67M) currently show poor semantic alignment, leading to misclassified products. Furthermore, at search time, 15.8% of the customer terms remain unmapped to our Search & Pub Knowledge Graph, inducing lower recall and precision. Solving these issues are a massive business opportunity.
While dense embeddings became the standard about semantic alignment, they often lack the lexical precision needed to fix these specific business gaps. This led us to explore an alternative: **Sparse Encoders**, a solution designed to bridge our massive product catalog with the explicitness of Knowledge Graphs to reclaim these lost data points.
We will walk through the evolution of our approach: from traditional term expansion (SPLADE) to a custom model with a key innovation: **using KG concepts themselves as tokens**. I will share the "behind-the-scenes" of our experimentation, including:
* **The Technical Pivot**: Why we moved from generative Seq2Seq models to Masked Language Modeling (MLM) and finally settled on a Distillation strategy.
* **Implementation Hardships**: Deep dive into the challenges of GPU optimization, custom DataCollators, and refining loss functions to enforce sparsity without losing semantic depth.
* **Real-world Results**: How we achieved 99%+ sparsity while maintaining alignment for complex product categories.

Beyond the code, this talk is a feedback loop on **data quality and model refinement**. Whether you are working on search, recommendation or data classification, you will gain practical insights into leveraging Sparse Encoders to build more robust and interpretable data products.

Speakers

Francois Gaillard

Digital Product Leader, ADEO Services

Engineering Manager at Adeo (Leroy Merlin, Bricoman) in AAAI team (Advanced Analysis and Artificial Intelligence) in Search & Publication domain. Last 10 years building and running AI powered search engines for e-commerce websites. PhD in computer science, specialized in Artificial... Read More →

Thursday May 28, 2026 12:00 - 12:30 CEST
4 - MEDIUM STAGE

Data / AI, Conference

Audience Advanced

12:15 CEST

Supply Chain operational performance through Agentic AI

Thursday May 28, 2026 12:15 - 12:45 CEST

3 - LAB

This presentation covers the evolution of Supply Chain automation at Leroy Merlin Ukraine. How the decommissioning of our old version of Pyxis led us to an AI Agent-enabled workflows in *almost* one step.

- The Challenge: Balancing legacy system deprecation and an upcoming SAP migration while trying to eliminate the manual resolution of micro-errors (e.g., typos, conflicting addresses) that traditional automation cannot catch.

- The Breakthrough: Proving that AI Agents succeed where rigid algorithms fail. By mimicking human reasoning to clean up unpredictable user inputs, this architecture is designed to directly reduce our Manual Touch Rate (Exception Rate) and drive up our First-Attempt Delivery Rate (FADR).

- The Demonstration: A practical look at our "start small, explore patterns, test assumptions" philosophy, showing how AI handles the tedious exceptions so humans can focus on delivering value.

This is a perfect opportunity to address smaller BUs like ourselves with a viable, scaled-down approach to Agentic AI. The goal: eliminate the countless micro-errors that constantly drain Supply Chain teams already operating at full capacity.

Speakers

Mykola Romanov

Product Owner, Leroy Merin Ukraine

Mykola Romanov
since 2024 ---> Product Owner of Supply Chain in LMUA
2021 - 2024 ---> leading BU`s innovative Products: eCommerce, Supply Chain, Payment & Loyalty
2019 - 2021 ---> Web Store Leader in LMUA (development and operations)
2014 - 2019 ---> Product Manager (Web Site Lau... Read More →

Thursday May 28, 2026 12:15 - 12:45 CEST
3 - LAB 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Beginner

14:00 CEST

Marketplace trust & safety catalog management solution: Keyword extraction & identification safety system (KEISS)

Thursday May 28, 2026 14:00 - 14:30 CEST

🤖 DATA/AI ARENA

This initiative introduces a robust, Python-based Generative AI solution engineered to safeguard a 18-million SKU marketplace from regulatory and brand safety risks. By automating the detection of non-compliant products and offensive content, this tool replaces inefficient manual keyword searches with a highly scalable, semantic validation engine.

The Challenge: The "Scale vs. Safety" Gap Following a massive catalog expansion in 2025, the organization experienced a surge in "Out of Scope" (OOS) listings. While illicit goods are centrally managed, two critical operational gaps emerged:

- Strategic Misalignment: Non-core products (e.g., smartphones, toys, or food) infiltrating a specialized DIY/Home Improvement marketplace.
- Brand Safety: Offensive content and hate speech hidden within product titles and descriptions.
- Inefficient Current State: Local Business Units (BUs) relied entirely on manual, unscalable keyword searches within the Mirakl platform to police the catalog.

The Solution: Hybrid AI Auditing Architecture To address these challenges, we developed a dual-approach architecture combining automated catalog screening with ad-hoc auditing tools:

1. Proactive Pipeline (Catalog-Wide Screening): A comprehensive, custom-built tool designed to validate the health and compliance of the entire country-level catalog, encompassing both published and pending (pre-publication) selection. This background process operates through a robust four-step automated workflow:
1.1 Initial Selection Filtering: Executes a broad sweep across the entire BU and supplier catalog, utilizing Regular Expressions (RegEx) and country-specific rules to rapidly identify and isolate products that are potentially Out of Scope (OOS).
1.2 Semantic Context Analysis: Leverages Vertex AI to conduct deep Natural Language Processing (NLP) on the isolated items. It evaluates full product titles and descriptions to understand their true semantic context (e.g., accurately distinguishing a prohibited "smartphone" from a permitted "smartphone repair tool").
1.3 OOS Classification: Categorizes the analyzed products with high precision, definitively flagging non-compliant items and separating them from legitimate, brand-safe merchandise.
1.4 Automated API File Generation: Seamlessly operationalizes the decisions by automatically generating ready-to-upload files (via API integration), eliminating manual data manipulation and enabling immediate catalog updates.

2. Reactive Pipeline (Targeted Auditing App): A custom Streamlit-based web application that empowers non-technical users to conduct deep, ad-hoc catalog audits through a four-step automated workflow:
2.1 AI Query Expansion: Leverages Gemini (Vertex AI) to dynamically expand a single seed keyword (e.g., "Barbie") into a comprehensive cluster of related terms (e.g., "dolls", "toys") to capture evasive listings.
2.2 Live Search Analysis: Audits front-end webpage results, capturing and evaluating exactly what the customer sees in real-time.
2.3 Semantic Validation Engine: A hybrid verification step combining BigQuery product data with Gemini’s Natural Language Processing (NLP) to ensure the product aligns with the BU’s strategic scope.
2.4 Operationalized Outputs: Automatically generates actionable files, including a Mirakl-formatted auto-rejection list for zero-GMV items, alongside a prioritized manual review queue for high-risk, revenue-generating products.

Impact & Current Results: The implementation has successfully transitioned the business from reactive compliance to proactive risk mitigation:
- Scale at Speed: Processes 274 SKUs per minute (approx. 3,300 SKUs in 12 minutes).
- Proven Output: To date, the system has screened 178,000+ products, analyzed 1,641 risk keywords, and successfully delisted 46,000+ non-compliant items.
- Precision & Revenue Protection: Maintains a False Positive Rate (FPR) of

Speakers

Mattia Mangia

Global Catalog Project Manager | Process & Automation, ADEO Services

Engineer. Designer. Maker. From Enterprise HQ buildings, Data Center, Hospitals Design to custom electronics and fine art, I bridge the gap between informatics and aesthetics. To me, the world is a series of complex projects, and my mission is to manage and deliver them with technical... Read More →

Maria Alexandra Ovalles

Senior Project Leader- EU Catalog Automation and Animation, ADEO

From Santo Domingo, Dominican Republic. Marketer at heart, now driven by data architecture and process optimization.Transitioned from branding and TV commercials to SQL, Python, and workflows.Proven track record in E-commerce giants: Glovo, Amazon, and LMES.

... Read More →

Javier Mena

Catalog Integation Team Leader, ADEO Services

- From San Salvador, El Salvador
- Catalog Integation Team Leader at AMS for +3 years, based in Madrid
- Previous ecommerce experience and background at ManoMano and Amazon
- Hobbies: Sports, Traveling, AI, Music

Thursday May 28, 2026 14:00 - 14:30 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Advanced

14:45 CEST

We will do them later - A Talk About LLM Evals

Thursday May 28, 2026 14:45 - 15:15 CEST

🤖 DATA/AI ARENA

Coding Agentic Systems is fun — it feels like magic!
Vibe-coding Agentic System is even more fun — double magic, hah!?

But quality evaluation of LLM generations? That part is usually… boring.
And if you’re not in the Python ecosystem, good luck finding a framework that actually works for you.
So eval tasks quietly sit in our backlogs, waiting for better days, the v2 release, or some future “we’ll fix it later.”

Drawing from real experience building production RAGs, agentic pipelines, and LLM-powered features across different stacks, I'll share the lessons learned the hard way — what broke, what worked, and what we wished we'd measured from day one.
The goal of this talk is simple: make evals understandable, and the creation process easy — with Coding Agents doing the heavy lifting alongside you. Practical tips, tricks, and a fresh perspective on making evaluation a natural part of developing LLM-powered applications.

Speakers

Nail Khusainov

Staff Engineer - ML/AI, ADEO Services

12 years cooking ML systems — occasionally burning them, but hey, sometimes they turn out great!

Thursday May 28, 2026 14:45 - 15:15 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Intermediate

15:00 CEST

MOTHERDUCK - Bullsh*t AI Is Over. Real analytics with just a prompt

Thursday May 28, 2026 15:00 - 15:45 CEST

1 - MAIN STAGE

We've spent years writing SQL, building Python pipelines, and waiting weeks for insights stuck behind a Jira ticket. What if anyone could go from a question to an answer with just a prompt?

Let's be honest, until late 2025, that was fiction. The demos looked great on stage. Production? Not so much. But something actually shifted. Models got genuinely good at structured output, MCP became the standard to connect LLMs to real tools, and DuckDB made it possible to query anything without infrastructure.

In this talk, I won't oversell. I'll build a full analytics workflow live — from raw data to insights using just natural language. First with a frontier model to show what's actually possible today with DuckDB and MotherDuck. Then with open-source models running entirely locally — no data leaving your network, no cloud dependency.

If you've been burned by AI demos that only work on slides, this one's for you. Expect a real demo, honest tradeoffs.

Speakers

Mehdi Ouazza

MotherDuck

I started my career in data 10+ years ago as a data engineer, working in large corporates like AXA setting up on-prem Spark clusters (yes, that old!) to tech unicorns building data platforms in the cloud at Klarna, Back Market, and Trade Republic.

Over the years, I found a passi... Read More →

Thursday May 28, 2026 15:00 - 15:45 CEST
1 - MAIN STAGE

Data / AI, Tools In Action

Audience Intermediate

15:30 CEST

Building a Conversational AI Platform Together: Governance, Architecture & REX

Thursday May 28, 2026 15:30 - 16:00 CEST

🤖 DATA/AI ARENA

How do you build a conversational AI platform that combines capabilities from multiple teams across an organization?

In this talk, we share how we designed, scaled, and governed a multi-assistant AI platform, from first integration to a reusable governance framework. Through concrete REX with Opus and Locus teams, we'll show how we went from case-by-case integrations to a clear model where any team can plug their AI capabilities into the platform.

We'll cover architecture decisions, integration patterns, and the prerequisites we now require before anything reaches production.

Speakers

Victor Bouy

ADEO Services

Software engineer turned GenAI integration specialist. 2 years building and shipping the conversational AI platform on the CCDP Care team, now part of the Data Accelerators team. Focused on delivering GenAI capabilities into customer-facing production systems.

Nadya Serbouti

Full-stack Engineer, Ext - ADEO Services

Fullstack engineer, 8 years of experience. Part of the CCDP Care team since March 2025. Working on the conversational AI platform across the full stack, from backend orchestration to frontend integration.

Thursday May 28, 2026 15:30 - 16:00 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Conference

Audience Advanced

16:15 CEST

Streaming to gold: How Obramax is scaling event-driven data pipelines with Kafka

Thursday May 28, 2026 16:15 - 16:45 CEST

🤖 DATA/AI ARENA

Designing a data architecture that can handle the **massive volume of a retail operation** requires more than just good tools; it requires **hardcore engineering rigor**.

At **Obramax**, we completely redesigned our data infrastructure, moving away from **isolated legacy databases** to a **unified, event-driven model** hosted on the **ADEO Group's DFDP platform**.

This session dives deep into the **engine room of our new Data Analytics platform**.

We will demonstrate our **aggressive automation approach** to building a **Medallion Architecture**.

You will see how we use **Apache Kafka** to capture **high-volume events** *(from SAP, VTEX, etc.)* and land them in our **Bronze layer**.

From there, we will showcase our strict **GitOps workflow using Terraform and dbt (Data Build Tool)** to automatically refine data into the **Silver layer**, and finally aggregate it into **business-ready Gold layer datasets**.

> Come see how we **automated the entire lifecycle of a Kafka event**
> from **system integration** to **BigQuery dashboard**
> with **zero-touch provisioning and maximum scalability**.

---

# 🎯 Key Takeaways

Participants will learn:

### ⚙️ Event Streaming Data Platform

A deep dive into building a **Medallion Architecture (Bronze, Silver, Gold)** using **event streams on Google BigQuery**.

---

### 🧱 Engineering Data Pipelines at Scale

How to combine:

- **Apache Kafka**
- **dbt (Data Build Tool)**
- **Terraform**

to build **bulletproof, scalable data pipelines** on the **ADEO DFDP platform**.

---

### 🔄 GitOps Data Infrastructure

Practical **GitOps strategies** for achieving:

- **zero-touch data infrastructure provisioning**
- reproducible data pipelines
- automated platform governance

---

# 🏗 Architecture Themes Covered

- Event-Driven Architecture
- Apache Kafka
- dbt (Data Build Tool)
- Terraform
- GitOps for Data Platforms
- Google BigQuery
- Data Lakehouse Pipelines

---

# 👨‍💻 Target Audience

This talk is ideal for:

- Data Engineers
- Platform Engineers
- Integration Engineers
- Data Infrastructure Teams
- DevOps / Platform Engineering Teams

---

# 🧭 Why This Talk Matters

Modern data platforms require **engineering discipline at scale**.

Retail environments generate **massive volumes of operational events**, and traditional batch pipelines cannot keep up with the **speed and reliability required for real-time analytics**.

By combining **event streaming, infrastructure-as-code, and GitOps practices**, Obramax built a **fully automated data platform** capable of transforming raw operational events into **business-ready analytical datasets**.

This session shares the **engineering patterns and automation strategies** that make this architecture scalable and maintainable.

---

# 🏢 Case Study

**Company:** Obramax

**Architecture Stack:**

- Apache Kafka
- ADEO DFDP Platform
- Google BigQuery
- dbt (Data Build Tool)
- Terraform
- GitOps Workflows
- Event-Driven Data Pipelines

Speakers

Samuel Xavier Da Silva

Architect, Obramax Brazil

# 👨‍💻 Samuel Xavier

Samuel Xavier is a **System Integration and Data Engineering Leader at Obramax**, part of the **ADEO Group**, where he leads initiatives focused on **event-driven architectures, large-scale data integration, and modern data platforms** in the retail e... Read More →

Thursday May 28, 2026 16:15 - 16:45 CEST
🤖 DATA/AI ARENA 135 Rue Sadi Carnot, Ronchin, France

Data / AI, Tools In Action

Audience Advanced