Learning - end-to-end RAG setup
Just wrapped up a PoC that deepened my understanding of RAG, with a strong focus on Convex and its ecosystem.
I explored how different Convex components fit together to build a real-time, end-to-end RAG setup — from document ingestion and vector search to agent-driven interactions.
Great learning experience combining modern RAG patterns with a highly productive full-stack setup.
Backend:
- @convex-dev/agent
- @convex-dev/rag
- @convex-dev/workpool
- OpenRouter
- LangChain
Frontend:
- React 19
- TanStack ecosystem
- shadcn/ui
- ai-elements by Vercel
GitHub: https://github.com/Krister-Johansson/rag
I explored how different Convex components fit together to build a real-time, end-to-end RAG setup — from document ingestion and vector search to agent-driven interactions.
Great learning experience combining modern RAG patterns with a highly productive full-stack setup.
Backend:
- @convex-dev/agent
- @convex-dev/rag
- @convex-dev/workpool
- OpenRouter
- LangChain
Frontend:
- React 19
- TanStack ecosystem
- shadcn/ui
- ai-elements by Vercel
GitHub: https://github.com/Krister-Johansson/rag
