Agenda

Look ma, no hands! Multimodal AI-Agents in the browser

Web Stage — 45 minutes Web LLM AI Agents

Move beyond chatbots and discover the true potential of on-device AI. In this talk, we'll explore how to build multimodal AI agents that understand voice commands, interpret images, maintain conversation context, and respond intelligently, all running entirely in the browser using modern web technologies like WebGPU and WebAssembly. We'll dive into practical patterns including AI agents, tool calling, and retrieval-augmented generation (RAG), demonstrating how LLMs can significantly enhance user experiences for real-world use cases. Best of all: these solutions run fully offline, require no cloud providers, and are privacy-preserving by default.Whether you're a developer looking to adopt cutting-edge techniques or a decision-maker evaluating new technologies, you'll leave with concrete examples of what's possible when AI meets the open web. Read More...

• Nico Martin [Open Source Machine Learning Engineer — Hugging Face]