Schedule

Bringing and Running AI Agents in the Browser

A growing number of web and JavaScript applications now rely on large language models (LLMs) and vector search/RAG (Retrieval-Augmented Generation) to deliver smarter, more intuitive user experiences. But the landscape is shifting once again. We’re entering the era of AI agents – a powerful paradigm that unifies multiple generative AI capabilities and streamlines how users interact with AI systems. By enabling agents to behave more like humans—making autonomous decisions, choosing tools, and solving tasks end-to-end—we move toward truly human-like, goal-driven automation. In this talk, we’ll begin with a concise overview of LLMs and RAG. From there, we’ll dive deeper into the architecture and capabilities of AI agents. We will learn how to use LLMs for targeted, specialized tasks, we will learn how tool-calling / function-calling works within agent workflows, and we will learn how agents independently select the right functions, APIs or data sources to ccomplish a user’s request efficiently and intelligently. We’ll also demonstrate how to build an AI agent that runs entirely in the browser, and compare that with building and running agents in the cloud. This session is designed for JavaScript developers, AI enthusiasts, and technical innovators who want to bring intelligent, autonomous capabilities into their web applications—and unlock the next generation of interactive user experiences.