Skip to main content

Document RAG Service

Local-first RAG microservice

Source Code

Overview

Privacy-preserving RAG stack for PDF/TXT ingestion, semantic retrieval, and local LLM generation using FastAPI, ChromaDB, SentenceTransformers, Ollama, and Docker.

Tech Stack

FastAPIFastAPI
ChromaDBChromaDB
SentenceTransformers
OllamaOllama
Next.jsNext.js
DockerDocker
Pytest

Key Highlights

  • Built ingestion, chunking, embedding, retrieval, and generation layers with async FastAPI routes.
  • Kept inference fully local through Ollama so documents never leave the user's machine.
  • Covered unit, integration, and E2E flows with 90%+ test coverage inside Docker.