LeetMentor (Local AI RAG)

LeetMentor is a privacy-first, local AI coding assistant that uses Retrieval-Augmented Generation (RAG) to provide grounded hints and mentorship based on actual successful and failed LeetCode submissions.

It completely decouples from cloud APIs by running Ollama locally for both text generation (e.g., llama3.1:8b) and text embeddings (nomic-embed-text), utilizing ChromaDB for blazing-fast vector storage.

Features

• Dual-RAG Memory: Instantly toggle your AI's brain between your own personal LeetCode submission history and a massive Global Dataset of thousands of highly-upvoted Python solutions.
• Sleek React Frontend: A premium, dark-mode (Vercel/Linear-inspired) UI built with React, TypeScript, and Vite.
• Real-Time Streaming: Tokens stream directly from the local LLM to the UI without blocking, utilizing Server-Sent Events (SSE).
• Context Inspection: A polished sliding right-panel allows you to natively inspect the exact retrieved vectors (the "sources") and telemetry metrics that the LLM used to formulate your hint.

Architecture

• Backend (/backend): FastAPI, Python 3.12, Uvicorn, ChromaDB, Ollama.
• Frontend (/frontend): React 18, TypeScript, Vite, lucide-react, react-markdown.

Getting Started

1. Prerequisites

• Python 3.11+
• Node.js & npm
• Ollama installed and running locally with the following models pulled:

  ollama pull llama3.1:8b
  ollama pull nomic-embed-text

2. Backend Setup

cd backend
python -m venv venv
source venv/bin/activate
pip install -r requirements.txt
cp .env.example .env  # Optional: tune any environment variables

3. Populating the Vector Database

You can evaluate your code against your own submissions or use a massive global parquet dataset.

# In the /backend directory (with venv activated)
python -m scripts.ingest_submissions  # Ingests your personal submissions.json
python -m scripts.ingest_global       # Ingests the global HuggingFace/Kaggle dataset

4. Running the Development Servers

Terminal 1 (Backend):

cd backend
source venv/bin/activate
uvicorn app.main:app --reload

Terminal 2 (Frontend):

cd frontend
npm install
npm run dev

The React interactive UI will be immediately available at http://localhost:5173.

License

This project is licensed under the MIT License - see the LICENSE file for details.