SupaChat – Conversational Analytics Platform

SupaChat is a full-stack conversational analytics application that allows users to query a PostgreSQL database using natural language and receive results as tables and visualizations. It is designed with production-grade DevOps practices including containerization, CI/CD, and monitoring.

🏗️ Architecture User (Browser) ↓ Next.js Frontend (Chat UI + Charts) ↓ FastAPI Backend (NL → MCP → SQL) ↓ Supabase PostgreSQL Database ↓ Monitoring (Prometheus + Grafana) 🔧 Tech Stack Frontend Next.js (App Router) React Axios Recharts Backend FastAPI Python MCP Query Translator (NL → JSON → SQL) Database Supabase (PostgreSQL) DevOps Docker Docker Compose Nginx (Reverse Proxy) GitHub Actions (CI/CD) Prometheus & Grafana (Monitoring) AI Tools Groq API (LLM inference) Prompt Engineering for SQL generation ⚙️ Setup Instructions

1. Clone Repo git clone https://github.com/copyfromabove/supachat.git cd supachat
2. Backend Setup cd backend pip install -r requirements.txt

Create .env:

GROQ_API_KEY=your_api_key DB_HOST=localhost DB_NAME=your_db DB_USER=your_user DB_PASSWORD=your_password

Run backend:

uvicorn main:app --reload 3. Frontend Setup cd frontend npm install

Create .env.local:

NEXT_PUBLIC_API_URL=http://localhost:5000

Run frontend:

npm run dev 🐳 Docker Setup Build & Run docker-compose up --build Services Frontend → http://localhost:3000 Backend → http://localhost:5000 Grafana → http://localhost:3001 Prometheus → http://localhost:9090 🚀 Deployment Steps Build Docker images Push to container registry (Docker Hub / ECR) Deploy on VM / Cloud (AWS / GCP) Configure Nginx reverse proxy Enable HTTPS (optional) 🔄 CI/CD Pipeline

Implemented using GitHub Actions

Workflow: Code Push → Build → Test → Docker Build → Deploy Features: Automated builds Linting & checks Docker image creation Deployment trigger 📊 Monitoring & Dashboards

Prometheus Collects backend metrics Tracks API latency & requests Grafana Visual dashboards System health monitoring Example Metrics: Request count Response time CPU & memory usage 🤖 MCP Query Translator (Core Feature) Flow: User Query → MCP JSON → SQL → Database Example:

Input:

Show top articles in last 30 days

MCP Output:

{ "action": "select", "table": "articles", "filters": [...], "order_by": "views DESC" }

SQL Generated:

SELECT * FROM articles ORDER BY views DESC; 📈 Features Natural language querying Chat-based UI Data visualization (charts + tables) Query history tracking Error handling & loading states Production-ready DevOps pipeline 🎥 Demo

👉 [Demo]

📂 Project Structure supachat/ ├── frontend/ # Next.js app ├── backend/ # FastAPI server ├── docker-compose.yml ├── nginx/ ├── monitoring/ └── .github/workflows/ 🔥 Future Improvements Authentication (JWT / OAuth) Role-based access control Query caching (Redis) Advanced analytics Multi-database support 👨‍💻 Author

Likhith Nagavelli

www.linkedin.com/in/likhith-nagavelli-1ab58b235 https://github.com/likhith777666?tab=repositories