Activité
1 k abonnés
-
Lionel Basdevant a partagé ceciDemain, j’aurai le plaisir de participer à une table ronde consacrée au rôle central de la data dans l’IA. Ce sera l’occasion d’échanger sur les transformations que l’IA apporte déjà au commerce et à la publicité digitale, sur l’importance d’une data de qualité pour créer de la valeur, et sur la manière dont Criteo accompagne les annonceurs et les marques dans cette évolution. Hâte de partager ces réflexions et d’en discuter ensemble!
-
Lionel Basdevant a republié ceciLionel Basdevant a republié ceciAs AI shopping experiences scale, recommendations are only as good as the data behind them. That’s why we’re introducing our Agentic Commerce Recommendation Service, built to help AI shopping experiences across LLMs and retailer sites deliver relevant product recommendations grounded in real-world shopping behavior. 🔹 In testing, our service delivered up to 60% higher recommendation relevancy compared to approaches based only on product descriptions. 🔹 And because it connects AI assistants directly to merchant inventory, the output is curated, transaction-ready, and aligned with real intent. Want to see how this brings trusted commerce intelligence into AI-driven shopping experiences? 🗞️ Read the full announcement here: hub.criteo.com/ukHwd
-
Lionel Basdevant a republié ceciLionel Basdevant a republié ceciDeep Learning models represent today’s cutting edge in performance recommendation systems. At Criteo, we’ve spent the last decade perfecting them — from scalable two-tower architectures to counterfactual evaluation and optimization with bandit feedback. These systems have become the backbone of measurable performance in digital commerce. Now, a new wave is emerging — that of LLM-based assistants, bringing reasoning, conversation, and trust into how people discover and decide. In my new article — The Future of Commerce Recommendation: Why Hybrid Systems Will Win — I outline how this convergence creates a new class of Recommendation Agents: 🔹 Grounded in data and performance, like today’s best recommenders 🔹Enriched with conversation and trust, powered by LLMs 🔹Built for scale and transparency across the entire commerce ecosystem This is where Criteo’s expertise and reach truly matter — connecting 17,000 advertisers, 230 retailers, and 4,000 brands through a shared, performance-optimized network. Read the full article: https://bit.ly/3J8reEg #AI #RecommenderSystems #HybridAI #LLMThe Future of Commerce Recommendation: Why Hybrid Systems Will Win | CriteoThe Future of Commerce Recommendation: Why Hybrid Systems Will Win | Criteo
-
Lionel Basdevant a publié ceciThe #GooglePrivacySandbox is a great example of collaborative work, in the open, showing that new, more private, ways of supporting digital advertising are possible. It's been used by Criteo to deliver millions of impressions a day. If everything wasn't perfect, that's an impressive milestone. Users and businesses are asking for more control and transparency over their data. Solutions with strong, demonstrable, privacy guarantees are the industry's future. Hanne Tuomisto-Inch Damien Faucherand Xavier Plantaz Michael Kleber Alex Cone Daniel Rojas
-
Lionel Basdevant a republié ceciLionel Basdevant a republié ceciGoogle’s latest Privacy Sandbox update isn’t the end of privacy progress—it’s a turning point. At NextRoll, we believe real innovation comes from shared momentum, not single-player solutions. In our response, NextRoll's VP of Data Science Engineering, Andrew Pascoe, outlines why this moment matters and how we’re continuing to build a more private, performant web—for advertisers and consumers alike. Read our POV. #PrivacySandbox #GoogleChrome #ThirdPartyCookies #OnlinePrivacy #PrivacyForwardAdvertisingNextRoll’s Response: Why Google’s Privacy Sandbox Pause Is a Call to InnovateNextRoll’s Response: Why Google’s Privacy Sandbox Pause Is a Call to Innovate
-
Lionel Basdevant a republié ceciLionel Basdevant a republié ceci🚀 Excited to share our latest insights on Private Measurement Systems! 🚀 As we dive deeper into discussions at the IAB Tech Lab Private Ads Systems TF (https://lnkd.in/eSeB63Qy) and the W3C on building a robust Private Measurement API (https://w3c.github.io/ppa/), we've published two insightful articles that explore the intricacies of ARA and its implications. 1. Coverage Gaps and Insights: This article delves into the coverage gaps within the Privacy Sandbox Attribution Reporting API and provides valuable insights to enhance its utility. 🔍📊 https://lnkd.in/eVF4UQym 2. Analysis of Summary Report Noise and Use Case Coverage: Our second article analyses the noise in summary reports and evaluates how well the current measurement APIs cover various use cases. 🔬 https://lnkd.in/emzadybn These articles aim to identify and address critical features, helping us create a Private Measurement API that is both useful and widely adopted in our thriving web ecosystem. Join the conversation and let's shape the future of private measurement together! 📈
-
Lionel Basdevant a republié ceciLionel Basdevant a republié ceciCriteo AI Lab is open-sourcing the largest real-world anonymised bidding dataset to push the boundaries towards future-proof private advertising systems, respecting user privacy while ensuring an economically viable open internet 🎉 . Are you working on privacy-enhancing technologies such as differential privacy and want to contribute to shaping the advertising industry? Do you want to benchmark performance brought by browser vendors proposals for bidding optimisation such as Chrome Privacy Sandbox? Then, CriteoPrivateAd dataset (https://lnkd.in/gR_tKEGU) is your playground! This dataset is available on Hugging Face and comes with a companion paper providing context, details and baselines (https://lnkd.in/gHyaMtzz)! As a leading AdTech company that drives commerce outcomes for media owners and marketers, Criteo is committed to evaluating proposals that might affect the way we will perform attribution, reporting and campaign optimisation in the future. We hope fostering research and industrial innovation on privacy-preserving machine learning applied to a major advertising use-case, namely bid optimisation under user signal loss. Diarmuid Gill Romain Lerallut Liva Ralaivola Lionel Basdevant Fatma Moalla François Griffon Nicolas Chrysanthos IAB Tech Lab Shailley Singh Damien Faucherand Nathalie Laneret European Tech Alliance Alliance Digitale Garrett Johnson Aurélien Bellet Roxana Geambasu Aymeric Dieuleveut Joaquin E.
-
Lionel Basdevant a republié ceciLionel Basdevant a republié ceciWho am I? I am a type of advertising that : ▶️ Is not viable at scale ▶️ Cuts publisher revenues in half ▶️ Is highly dependent on the context However AI and commerce data can bring me a lot of value Thank you Lionel Basdevant for your lightening talk today at the IAB Europe horizon conference in Brussels on whether contextual advertising is economically viable! Pierre Devoize Ninon Vagner Romain Lerallut Loïc Rivière Inés T. Enrico Girotto Robin de Wouters Carolin Wehrhahn Laurent T. Emma A. Daniel Knapp Marieke Buis Florian Herzing Anca Caruntu Thaima Samman
-
Lionel Basdevant a partagé ceciHappy to be speaking at IAB Europe Advertising Horizons, next week in Brussels. 👉 You can still register for the event: https://lnkd.in/epKp-fTU
-
Lionel Basdevant a aimé ceciLionel Basdevant a aimé ceciAds & Media Day organisé ce matin chez Criteo par La villa numeris et THE GOOD ADVERTISING PROJECT On parle Data, IA, contrôle des écrans, collecte des signaux, avec les acteurs de la chaîne de valeur ajoutée : médias, techs, enseignes de distribution,… Tous développent leurs agents IA pour une exploitation de la data encore plus pertinente et instantanée, et tous deviennent des points de collecte des signaux. Mais cela pose la question de la mise à jour nécessaire du cadre réglementaire. Merci david lacombled et Loïc Rivière pour votre confiance. Merci à mes voisins de table ronde pour leur input : Priya Saint Olive, Francois LOVITON, Vincent PETILLO, Lionel Basdevant et MICHEL JUVILLIER pour la qualité de ton animation! Et sur le sujet de la régulation : Julie Carel, Nathalie Laneret, Mélanie Defoort, Marine C. Gossa, et Laureline L'Honnen-Frossard
-
Lionel Basdevant a aimé ceciLionel Basdevant a aimé ceciBest side quest of the summer - seeing my maps in the exhibit at the French National Library - right alongside the original drafts of middle earth. I spent so long immersed in Westeros, Essos, and the Lands of Ice and Fire when I created these - it’s wonderful to see others immerse themselves in the exhibit.
-
Lionel Basdevant a aimé ceciLionel Basdevant a aimé ceciCriteo shared a Press Release this week on our integration with OpenAI and where we stand on the ChatGPT ads pilot. I sat down with Lionel Basdevant, Product Director for Agentic Commerce, to unpack the key insights from and what it means in practice for marketers today. In just 5 minutes, we cover: 💎 #Pilot status: 2000 brands live, 7 countries activated. Apparel, home furnishings, consumer electronics, automotive and beauty have emerged as the leading categories running ads in ChatGPT through Criteo 💎 #CPCbidding and what performance marketers should watch as optimization evolves beyond impressions 💎 #PromptSmartAds, adapting to the intent and context of a conversation by leveraging #CommerceIntelligence 💎 #ConversionModeling and measurement approaches designed for conversational AI experiences 💎 #Insights to help brands better understand how shoppers discover and evaluate products in AI-driven environments No breaking news. No confidential roadmap. (Rest assured, marketing team 😉) Just a practical snapshot of where AI-native advertising stands today, what we're seeing in the pilot, and what marketers should start paying attention to now. The question is no longer whether consumers will discover products through AI. It's how quickly marketers adapt. Criteo's full press release: https://lnkd.in/edannHVy We're only at the beginning of this journey 🚀 Connect with our teams around the world to explore what's next in AI-native advertising 🌎 Parker Bohlen Marc Fischli Grace Fromm Szi-Wei Lo Matt Hurle Joaquin Gallardo Mattia Rovelli Marthe Seguin Laure Sauteraud David LAGARDERE Marc Feu
-
Lionel Basdevant a aimé ceciLionel Basdevant a aimé ceciHave you noticed how the conversation around #AI in Product so often jumps straight to tools, shortcuts, and hype? 🤔 For me, the more interesting question is deeper: what does Product craft actually look like in an AI-native world? And just as importantly, what parts of it remain timeless? That’s exactly why we’re launching our first Product x AI meetup in Paris, to keep building community with product people and create space for the conversations that really matter right now. The event is centered on how companies are navigating product craft in an AI world, with room for genuine networking among people excited about both AI and Product. 🤝 We’ll be hosting it at Criteo, with an agenda that includes welcoming remarks, a session on building an AI-fluent culture, a panel on what Product craft looks like in an AI-native world, and time to continue the conversation informally afterward. 🍸 Speakers and panelists include Julie Robles from PayFit, Sara Skrmetti, and Jérome Caubert from Criteo. 🎤 This is possible thanks to Swati Singh & Raluca Mateas with the priceless support from Todd Parsons & Diarmuid Gill 👏 I’m especially looking forward to exchanging on questions like these: what has really changed in discovery, prioritization, and delivery? Where are teams overestimating AI today, and where are they still underestimating it? How should product work differently with engineering, design, and data science now? And what skills will matter most for product people over the next few years? If these are the kinds of conversations you’ve been wanting to have with peers, I’d love to see you there 🫶 Agenda + registration 👉 https://luma.com/gzoluf0q #ProductManagement #AI #TechCommunity
-
Lionel Basdevant a aimé ceciLionel Basdevant a aimé ceciExcited that Criteo, as the first adtech partner to OpenAI, is now live in the UK with ads in ChatGPT! 🇬🇧 We’re opening up a new environment for brands and retailers who want a front-row seat to what advertising looks like inside AI-powered conversations - and to help shape how we make it more relevant for consumers than ever before. I took the opportunity today for a rooftop conversation with Basile Leparmentier, one of the many people in our R&D team who helped bring this launch to life. What makes this space so compelling is the shift in how paid media create value. We're heading to a world where: 💎 Relevance becomes the primary driver: Every impression is earned in context, not who paid the most for a generic keyword 💎 Intent is explicit: Consumers are no longer just browsing signals, they are expressing needs directly, in real time, through conversations 💎 Performance becomes more about helping discovery: Enabling consumers to find new brands and products at the right moment We’re still early, and we’re going to learn a lot: How this new conversational environment drives net-new demand, how it reshapes discovery, and how we can surface richer intent and product insights back to brands. If you want a front-row seat and want to help shape what comes next, reach out to our Performance Media UK teams lead by Joaquin Gallardo and Mattia Rovelli to go live! #performancemedia #advertising #innovation
-
Lionel Basdevant a aimé ceciLionel Basdevant a aimé ceci🗽Quelle semaine incroyable à New York ! Je repars de Big Apple avec des étoiles plein les yeux et une immense gratitude pour la chance d'avoir vécu cette expérience unique au cœur de l'innovation... 🤝 J1 & 2 Enterprise CAB (Customer Advisory Board) : 2 journées intenses de co-construction et d'échanges stratégiques avec nos plus grands clients Enterprise. Leurs retours sont notre plus grande force! 🚀 J3 Navigate US : Le point d’orgue de la semaine à New York ! Une journée d’exception axée sur le futur du voyage d'affaires, de la gestion des dépenses (T&E) et de la révolution de l'IA (grâce à notre techno TravelClaw!) 🎯 J4 & 5 Mid-Market CAB : 2 jours captivants dédiés à l'agilité et aux ambitions de nos clients Mid-Market. Des discussions passionnantes pr aligner notre feuille de route sur leurs besoins de croissance... 👉 Ce que je retiens de cette semaine: Des conférences ultra-inspirantes, des annonces d'innovations majeures, des investisseurs et des médias captivés, et des sponsors ravis. Mais surtout : des clients hyper satisfaits et des prospects conquis par notre vision ��� 👏 BRAVO à toute l'Équipe Marketing pour l'organisation, c'était amazing! ✨ Un immense merci à ma manager Kim Hamer pour cette opportunité et sa confiance. Revenir de New York avec une telle énergie et autant d'idées, c'est le meilleur des moteurs ! 🇫🇷 Prochaine étape ? On ramène cette énergie à la maison... RDV au prochain Navigate à Paris en octobre ! Qui sera de la partie ? —---------------------- 🗽 What an incredible week in New York City! I’m leaving the Big Apple inspired, energized, and deeply grateful for the opportunity to have experienced such a unique event at the very heart of innovation. 🤝 Days 1 & 2 Enterprise CAB (Customer Advisory Board): Two intense days of co-creation and strategic exchanges with our largest Enterprise customers. Their feedback is our greatest strength. 🚀 Day 3 Navigate US: The absolute highlight of the week in NYC! An exceptional day focused on the future of business travel, T&E management, and the AI revolution (powered by our TravelClaw technology!). 🎯Days 4 & 5 Mid-Market CAB: 2 captivating days dedicated to the agility and ambitions of our Mid-Market clients. Thrilling discussions to align our roadmap with their growth needs. 👉 My key takeaways from this week: Ultra-inspiring keynotes, major innovation announcements, captivated investors and media, and thrilled sponsors. But above all: hyper-satisfied clients and prospects fully won over by our vision! 👏 CONGRATS to the entire Marketing Team for the organization, it was truly amazing! ✨ A massive thank you to my manager Kim Hamer for this great opportunity and her trust. Coming back from New York with this much energy and so many ideas is the ultimate driving force! 🇫🇷 Next stop? Bringing that energy back home... See you at the next Navigate in Paris this October! Who's joining? #Networking #Innovation #CustomerSuccess #Navigate2026 #NewYork #TravelTech
-
Lionel Basdevant a aimé ceciLionel Basdevant a aimé ceciRavi d'être à nouveau reconnu cette année en Public law par @Chambers and Partners. Cette distinction reflète avant tout l'expertise de l'équipe et la confiance de nos clients sur des enjeux stratégiques : commande publique, domanialité publique, projets d’infrastructures complexes, financement des collectivités territoriales et droit des aides d’État. Un immense merci à nos clients pour leur fidélité et à toute l'équipe pour son implication. * Delighted to be recognized once again this year in Public Law by @Chambers and Partners. This distinction above all reflects our team's expertise and our clients' trust in strategic matters: public procurement, public domain, complex infrastructure projects, local authority financing, and state aid law. A huge thank you to our clients for their loyalty and to the entire team for their commitment! #ChambersFrance2026 #PublicLaw #JoffeAssocies
-
Lionel Basdevant a aimé ceciLionel Basdevant a aimé ceciAs AI shopping experiences scale, recommendations are only as good as the data behind them. That’s why we’re introducing our Agentic Commerce Recommendation Service, built to help AI shopping experiences across LLMs and retailer sites deliver relevant product recommendations grounded in real-world shopping behavior. 🔹 In testing, our service delivered up to 60% higher recommendation relevancy compared to approaches based only on product descriptions. 🔹 And because it connects AI assistants directly to merchant inventory, the output is curated, transaction-ready, and aligned with real intent. Want to see how this brings trusted commerce intelligence into AI-driven shopping experiences? 🗞️ Read the full announcement here: hub.criteo.com/ukHwd
Expérience et formation
Licences et certifications
-
Structuring Machine Learning Projects by deeplearning.ai on Coursera
Coursera Course Certificates
Délivrance le -
Neural Networks and Deep Learning by deeplearning.ai on Coursera
Coursera Course Certificates
Délivrance le
Langues
-
French
-
-
English
-
Recommandations reçues
1 personne a recommandé Lionel
Inscrivez-vous pour y accéderAutres profils similaires
Découvrir plus de posts
-
Ben Vicca
I had the opportunity to share my perspective in an RTL Info feature on the growing value, and risks of personal data. Every click, every purchase, every preference we share online adds to a digital profile that’s extremely valuable for companies, and sometimes, unfortunately, for fraudsters too. As I mentioned during the interview: "If I receive messages that truly match my interests and needs, of course I'll read them, that's what gives data its value." But this value comes with responsibility. The future of digital business depends on trust, transparency, and respect for privacy. And as individuals, we also play a part. Every time we share our data, we make a choice. It’s worth pausing to understand how it will be used, and whether the benefit we get in return feels right. We’re fortunate to live in a region where data protection is not an afterthought but a fundamental right. GDPR exists for a reason, to keep control and sovereignty where it belongs: with us 😉. Knowing where your data lives, and who has access to it, is part of building a digital future we can all believe in. 📺 You can watch the full interview here: https://lnkd.in/ejNMR3HY #DataPrivacy #DigitalTrust #Inetum #CyberSecurity #AI #EthicalTech #InetumWay #DigitalImpact #ForTheGoodOfAll
2 commentaires -
Dan Adams
The inconvenient truth about AI? Real data is expensive. Everyone wants the outcomes – smarter predictions, faster insights, competitive advantage – but not everyone is willing to make the necessary investment in trusted, high-quality data. It’s become easier and more acceptable to pour money into GPUs and the latest models than to face the reality that improving inputs (and doing the work that entails) will always improve outcomes. The truth is, data integrity isn’t free. It takes resources to ensure datasets are accurate, complete, traceable, and bias-aware. But it’s also the investment with the highest return. When you get the data right, AI models not only perform better – they produce outcomes you can actually trust. Skipping this step is a slippery slope. Failed pilots, hallucinations, wasted compute, and reputational risk all add up to costs far higher than what it would’ve taken to build on a foundation of clean, contextual data. The companies that will win with AI aren’t just the ones running the biggest models. They’re the ones willing to invest in the integrity of the data beneath them. So before investing in the next model, ask: do you trust the data you’re feeding it? Because if you don’t, the output won’t be worth much. #AI #DataIntegrity #DataQuality #MachineLearning #TrustedData #AITrust #ResponsibleAI
1 commentaire