← back
Building Reactive AI Apps: Matt Welsh
Takeaway
AI.JSX brings React's composability to LLM programs so frontend devs can stream and compose model calls as nested components.
Summary
- Matt Welsh (Fixie) introduces AI.JSX, an open-source TypeScript framework that brings React-style component composition to LLM apps for frontend/full-stack devs.
- JSX components render asynchronously and in parallel to the LLM rather than the DOM, enabling concurrent streaming chat completions inside a tree (make-story → character/setting/plot).
- Built-in support for RAG (DocsQA over indexed corpora), tool calling via UseTools, multi-provider (Anthropic/OpenAI), and a floating Fixie embed React component for one-line deploy.
- Composition unlocks patterns like wrapping a component in a KidSafe component to enforce output rewrites; ~10 lines of JSX yields a full RAG app.
frameworktypescriptrag
Original description
AI.JSX is like React for LLMs -- it lets you build powerful, conversational AI apps using the power of TypeScript and JSX. Building LLM-powered apps is not like building other kinds of software. Not only can the AI model allow you to converse in natural language, but we can also use the LLM itself to replace a lot of the conventional code needed for processing data, interfacing with external APIs, and more. AI.JSX embodies this philosophy and makes it easy to build sophisticated AI experiences without a lot of code. It's also fully integrated with the Fixie platform, making deployment and management a breeze. Recorded live in San Francisco at the AI Engineer Summit 2023. See the full schedule of talks at https://ai.engineer/summit/schedule & join us at the AI Engineer World's Fair in 2024! Get your tickets today at https://ai.engineer/worlds-fair About Matt Welsh Matt Welsh is the Chief Architect and Co-founder of Fixie.ai, a Seattle-based startup developing a new computational platform with AI at the core. He was previously head of engineering at OctoML, a software engineer at Apple and Xnor.ai, engineering director at Google, and a Professor of Computer Science at Harvard University. He holds a PhD from UC Berkeley.