Maester — RAG Chat Application

A document-based AI assistant. Upload your files, ask questions, get answers grounded in your documents.

Features

• Document Q&A — Upload PDFs, Word, Excel, TXT, CSV, JSON and chat with them
• User Accounts — Register/login with username & password; each user's data is fully isolated
• Chat History — Conversations saved and resumable across sessions
• Conversation Memory — Summary buffer: LLM retains context from earlier in the conversation
• File Selection — Choose which uploaded files to use as context per query
• File Download — Download originally uploaded files directly from S3
• Source Citations — Every answer shows which document chunks were used with confidence scores
• Intent Detection — Chitchat handled directly; document queries go through RAG pipeline

Tech Stack

Layer	Tech
Frontend	React + Vite + Tailwind
Backend	FastAPI + Python
Vector DB	MongoDB Atlas Vector Search
Embeddings	all-MiniLM-L6-v2 (sentence-transformers)
LLM	Llama 3.1 via Groq
Storage	AWS S3
Deployment	AWS Lambda (Docker) + Vercel
Auth	JWT (15-day tokens) + bcrypt

Local Setup

Backend

cd backend
pip install -r requirements.txt
python main.py

Frontend

cd frontend
npm install
npm run dev

Set VITE_API_URL=http://127.0.0.1:8000/api in frontend/.env.

Environment Variables

GROQ_API_KEY=
MONGODB_URI=
JWT_SECRET=
AWS_REGION=ap-south-1
S3_BUCKET=

Deploy

cd backend
.\deploy-lambda.ps1