Skip to content

jwteeba/chroma_knowledge_search

Repository files navigation

📚 Chroma Knowledge Search — RAG System

AI-powered semantic document search platform enabling context-aware Q&A across uploaded PDFs/DOCX/TXT documents.

Uses

  • 🧠 OpenAI — embeddings + LLM responses + moderation
  • 🔎 ChromaDB — vector storage & retrieval
  • ⚙️ FastAPI — secure backend RAG service (Deploy via Render)
  • 🎛️ Streamlit — interactive upload + question UI
  • 🔐 API Key Authentication
  • 🗄️ SQLite metadata DB
  • 🐳 Docker Compose — fully containerized
  • 📦 Poetry — dependency management

🚀 Features

Capability Status
Upload PDF, DOCX, TXT
Automatic text extraction
Chunking with embeddings
Secure API-Key isolation
Contextual answers (RAG)
Safe content moderation
Storage in ChromaDB
Metadata persistence
Multi-doc querying
Containerized deployment
Fully local execution

▶️ Run the App

docker compose up --build
Service URL
✅ App UI http://localhost:8501
✅ Backend Docs http://localhost:8000/docs
✅ Backend Health http://localhost:8000/health
✅ ChromaDB REST http://localhost:8001/api/v2/heartbeat

About

Retrieval-Augmented Generation (RAG) system combining ChromaDB + OpenAI embeddings for contextual document Q&A across PDFs and knowledge bases. Backend powered by FastAPI, frontend in Streamlit.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors