Skip to content

README.md

Latest commit

 

History

History
88 lines (65 loc) · 1.55 KB

README.md

File metadata and controls

88 lines (65 loc) · 1.55 KB

Codebase AI Assistant

An intelligent AI-powered codebase assistant that allows users to upload a repository and ask questions about the code.
Built using RAG (Retrieval-Augmented Generation) with Qdrant + LLM + FastAPI + React.

Live Demo

Features

  • Upload GitHub repository or local codebase
  • Semantic search using vector embeddings
  • Ask questions about your code
  • Get file-based explanations
  • Supports multiple programming languages
  • ⚡ Fast retrieval using Qdrant vector database

Architecture

User Question
↓
Frontend (React / Vercel)
↓
FastAPI Backend (Hugging Face)
↓
Qdrant Vector DB
↓
Relevant Code Context
↓
LLM (Groq)
↓
Answer

Tech Stack

🔹 Backend

  • FastAPI
  • LangChain
  • Qdrant Vector Database
  • Groq LLM (LLaMA 3)
  • HuggingFace Embeddings

🔹 Frontend

  • React / Next.js
  • Tailwind CSS
  • Vercel Deployment

⚙️ How It Works

  1. Upload a repository
  2. Files are filtered and processed
  3. Converted into embeddings
  4. Stored in Qdrant
  5. User asks a question
  6. Relevant code is retrieved
  7. LLM generates answer based on context

Installation (Backend)

git clone https://github.com/kumar-kiran-24/Codebase-AI-Assistant.git
cd Codebase-AI-Assistant

pip install -r requirements.txt

Environment Variables(.env)

QDRANT_URL=your_qdrant_url
QDRANT_API_KEY=your_api_key
GROQ_API=your_groq_api_key