AI & Machine Learning
RAG Factory
Personal / Open Source
Ongoing
Creator / Maintainer
AI Infrastructure & LLM
Tech Stack
RAG
pgvector
ChromaDB
Mistral
Ollama
OpenAI API
React
FastAPI
Summary
Universal RAG engine converting any input source into a queryable vector knowledge base across multiple LLM providers.
What I Built
-
Built a universal RAG engine that ingests any input format — text, ODF, Git repos, S3 buckets, URLs — into a unified pipeline.
-
Implemented vector storage and retrieval across pgvector and ChromaDB backends.
-
Integrated multiple LLM providers (Mistral, Ollama, OpenAI) behind a common interface.
-
Built a React UI and FastAPI backend for end-to-end document ingestion and querying.