LEARN FROM THE BEST
ARMAND JOULIN
🔎 tech talk
Improving Open Language Models at a Practical Size
Armand is a Research Director at Google DeepMind, in charge of the open version of Gemini called Gemma. Prior to this, he was a Research Director at Meta in charge of EMEA, where he supervised several key open projects such as LLaMA, DINO or fasttext.
ROMAIN HUEt
🔎 tech talk
Building with OpenAI: What's Ahead
Romain currently leads Developer Experience at OpenAI, inspiring and supporting founders and builders to integrate AI into their applications, and directing the creation of elegant and powerful tools for all developers. Previously, Romain spent 5 years at Stripe, leading product management for the developer platform and overseeing global developer relations. Before Stripe, he helped with the relaunch of Twitter’s developer platform and co-founded Jolicloud in Paris, where he developed a cloud-based OS.
PIERRE STOCK
🔎 tech talk
The Future of Edge Agents
Pierre Stock is a Research Scientist and first employee at Mistral AI, mainly focusing on Code and Edge models and their efficient deployment to production. Before that, he was a Research Scientist at Meta and developed privacy-related machine learning techniques such as Federated Learning and Differential Privacy specifically for LLMs. Pierre Stock was a resident PhD student at Facebook AI Research (FAIR, CIFRE program) for 3 years.
GAEL VAROQUAUX
Skrub: Less data wrangling, more machine learning
Gaël Varoquaux is a research director working on data science at Inria where he leads the Soda team. He is also co-founder and scientific advisor of Probabl. His research covers fundamentals of AI, statistical learning, NLP, causal inference, as well as applications to health. He also co-funded scikit-learn, one of the reference machine-learning toolboxes, and helped build various central tools for data analysis in Python.
🔎 tech talk
INES MONTANI
🔎 tech talk
Reality is not an end-to-end prediction problem: Applied NLP in the age of Generative AI
Ines Montani is a developer specializing in tools for AI and NLP technology. She’s the co-founder and CEO of Explosion and a core developer of spaCy, a popular open-source library for Natural Language Processing in Python, and Prodigy, a modern annotation tool for creating training data for machine learning models.
STANISLAS POLU
🔎 tech talk
LLMs reasoning and agentic capabilities over time
After studying at Polytechnique and Stanford, Stanislas started a company in France that eventually got acquired by Stripe in 2015. He
spent 5 years at Stripe working on the API and infrastructure around non-card payment methods. After Stripe he did research for 3 years at OpenAI on large language models mathematical reasoning capabilities.
He's now co-founding Dust, building the company OS for smart teams.
NEIL ZEGHIDOUR
Multimodal Language models
Neil is co-founder and Chief Modeling Officer of the Kyutai non-profit research lab. He was previously at Google DeepMind, where he led a team on generative audio, with contributions including Google’s first text-to-music API, and the first neural audio codec that outperforms general-purpose audio codecs. Before that, Neil spent 3 years at Facebook AI Research, working on automatic speech recognition and audio understanding.
🔎 tech talk
ELIOT ANDRES
Training a foundation image model and deploying it to millions of users
Eliot Andres co-founded Photoroom to democratize image editing through AI. As CTO, he leverages 9 years of experience in deep learning and computer vision to put state-of-the-art algorithms into the hands of millions of users. He's currently focusing on training a GenAI foundation model for image generation and making capabilities available to other companies through a powerful API.
🔎 tech talk
SRI AMBATI
🔎 tech talk
Democratizing AI with Open Source Multi-Agent Systems: Advancing the Future of Workflow Automation
Sri founded H2O.ai in 2012 with a mission to democratize AI for anyone, anywhere, creating a movement of the world’s top data scientists, physicists, academics and technologists at more than 20,000 organizations worldwide. Most recently, Sri led the initiative o2forindia.org, sourcing O2 concentrators for more than 200 public health organizations in Tier 2 cities and rural communities in India during the Delta wave of the COVID-19 pandemic, helping to save thousands of lives.
MERyEM ARIK
🔎 tech talk
Getting LLMs to do what you want: Output Controllers
Meryem co-founded TitanML to create a seamless and secure infrastructure for enterprise LLM deployments. She studied Theoretical Physics and Philosophy at the University of Oxford and is a Forbes 30 Under 30 honoree. Beyond her daily role, Meryem is dedicated to envisioning an ethical AI future and enhancing diversity within the AI space.
STEEVE MORIN
🔎 tech talk
ZML: ML framework for Zig
Steeve Morin is a software engineer and co-founder of ZML, a high performance AI inference stack that's built for production, delivering peak performance on a wide range of chips. He is the author of LegiGPT, the law AI assistant, boot2docker and used to be VP Engineering at Zenly, acquired by Snap Inc.
DR. LAURE SEUGÉ AND ARTHUR TALPAERT
🤝 duo talk
More empathy and health data protection in AI: announcing a primary care revolution.
Doctor Laure Seugé is a pediatrician specialized in nephrology and rheumatology. She sees patients at the Children's Institute, Marcel Sembat Clinic in Boulogne Billancourt, and works part-time at Necker-Enfants Malades Hospital in pediatric nephrology (Paris). She is also a Medical Expert at Doctolib to advise the development teams in the product innovation and co-building.
Arthur is Head of AI Product, Consultation Assistant at Doctolib. He manages the design, build and launch of this product in 2024, and advises the Executive Committee on Doctolib's AI and data transformation. Arthur launched and managed tens of AI and data products, incl. at McKinsey Digital and at his own startup, Familiaris MedTech. Before that, Arthur was the Lead Data Scientist for industrial and MarTech products. He holds a PhD in Applied Mathematics, pertaining to the nuclear industry and national labs.
YANN LÉGER
🔎 tech talk
Serverless Inferencing: an infrastructure point of view
Yann Leger is co-founder of Koyeb, a serverless platform for AI workloads, and spent the last 12 years building large-scale cloud service providers from scratch. Passionate about cloud computing, he has a deep understanding of the underlying infrastructure, from data centers to the software stack running on hypervisors. After building Scaleway, originally with bare metal ARM servers, Yann decided to go serverless with alternative chips for AI.
MARJOLAINE GRONDIN
🧠 soft talk
AI is your personal co-founder
Marjolaine Grondin is the co-founder of Jam, the pioneering French-speaking chatbot recently acquired by June Marketing. After studying at Sciences Po, she discovered her passion for entrepreneurship at UC Berkeley and completed her education at HEC Paris. Featured in Forbes' 30 Under 30 and MIT's Top Innovators Under 35, Marjolaine is also the first female entrepreneur to speak at Meta's F8 conference.
HOSTED BY:
CLAIRE LEBARZ
🎙️ MASTER OF CEREMONY
Claire Lebarz is Chief Data and AI Officer at Malt. Prior to that, she led several teams at Airbnb, in charge of measuring and improving online and offline user experience with data and machine learning. She spent over 10 years in San Francisco, developing products in scale-ups and has worked with dozens of startups as an advisor or business angel.
ori pekelman
⚡ Lightning talk
Talk: AI Model Deployment Strategies: Balancing Efficiency and Environmental Impact
Ori is a Co-Founder of Platform.sh, a cutting-edge cloud hosting and development platform. He has been a steadfast advocate for Open Source, Privacy and reducing the Environmental Impact of computing. Platform.sh is recognized as a bcorp, with independent rating agencies EcoVadis and Greenly awarding it gold medals, ranking the company among the top 5% of all companies in all industries the agencies assess.
audrey roy greenfeld
⚡ Lightning talk
Talk: Rethinking AI: Building What Matters, Not Just What's Possible
Audrey is a Research and Development staff member at Answer.AI. She is co-author of the Two Scoops of Django book series, and the Cookiecutter project templating system. Previously, she was the cofounder and first president of PyLadies, and has worked in engineering and operations leadership roles in nonprofit and insuretech. Her perspective is further informed by early exposure to computer vision and robotics projects at MIT (2002-2005).
RAFFI sarkissian
⚡ Lightning talk
Talk: AI is expensive - Pricing and billing without engineering nightmares
Raffi Sarkissian co-founded Lago, used by cutting-edge companies such as Mistral, Groq, Together.ai, thousands of engineers, and backed by FirstMark, Y Combinator, Signal Fire, Meghan Gill (MongoDB), Clément Delangue (Hugging Face). Before co-founding Lago, Raffi was one of the first employees of Qonto.com, the $5B Brex for Europe.
Grigorij Dudnik
⚡ Lightning talk
Talk: Multi-agent collaboration in practice: make AI code an entire application for you
Grigorij is an AI developer working with AI agents. Author of Clean Coder - a multi-agent framework for autonomous code writing, which is completely open-source. He used Clean Coder to write most of the code in his startup takzyli.pl, where Grigorij is CTO and co-founder.