RAG Engine

Overview

A production-ready retrieval-augmented generation system with pluggable LLM providers, vector search, and a full-stack app for managing the knowledge base.

Highlights

• Engineered production-ready RAG system supporting 3 LLM providers (OpenAI, Gemini, Ollama) with intelligent vector search using Chroma and Qdrant, reducing query response time by 40% through caching optimization
• Built full-stack application with RESTful APIs, admin dashboard, and CLI tools for document ingestion and model orchestration, enabling knowledge-base management for 100+ documents with Docker deployment

Built with

• FastAPI
• Next.js
• React
• Python
• TypeScript
• Docker
• Chroma
• Qdrant
• OpenAI
• Gemini
• Ollama

Next project