AI Data Analysis Agent

An AI-powered chat application for data analysis. Upload CSV or Excel files and get intelligent insights.

Online Demo

test case

Graph Visualization

Features

• 📊 Upload CSV and Excel files
• 🤖 AI-driven data analysis and querying
• 💬 Conversational interface
• 📈 Data visualization (charts)
• 💾 Conversation history
• 📱 Responsive design with mobile support

Tech Stack

Backend

• FastAPI - Web framework
• LangChain - AI agent framework
• Pandas - data processing
• OpenAI GPT - language model
• Python 3.8+

Frontend

• Next.js 14 - React framework
• TypeScript - type safety
• Tailwind CSS - styling framework
• Recharts - data visualization

Deployment

Backend Deployment (Render)

1. Ensure backend/requirements.txt includes all dependencies:

fastapi==0.115.11
uvicorn==0.34.1
python-multipart==0.0.20
pandas==2.2.3
tabulate==0.9.0
langchain==0.3.25
langchain-openai==0.2.14
langchain-experimental==0.3.4
langchain-core==0.3.59
langgraph==0.3.31
langgraph-checkpoint==2.0.24
openai==1.75.0
python-dotenv==1.1.0
pydantic==2.11.3
typing-extensions==4.13.2
openpyxl==3.1.5
requests==2.32.3

2. Create a new Web Service on Render
3. Connect the GitHub repository
4. Set the build command: pip install -r backend/requirements.txt
5. Set the start command: cd backend && python main.py
6. Add an environment variable:
   • OPENAI_API_KEY: your OpenAI API key

Frontend Deployment (Vercel)

1. Import the GitHub repository in Vercel
2. Set the root directory to frontend
3. Add an environment variable:
   • NEXT_PUBLIC_API_URL: backend API URL (e.g. https://your-backend.onrender.com)

Local Development

Backend Setup

1. Enter the backend directory:

cd backend

2. Install dependencies:

pip install -r requirements.txt

3. Create a .env file:

OPENAI_API_KEY=your_openai_api_key_here

4. Start the server:

python main.py

Frontend Setup

1. Enter the frontend directory:

cd frontend

2. Install dependencies:

npm install

3. Create a .env.local file:

NEXT_PUBLIC_API_URL=http://localhost:8000

4. Start the development server:

npm run dev

Usage

1. Open the frontend application
2. Click "Add files" to upload CSV or Excel files
3. Type your question in the chat box, for example:
   • "Analyze the basic statistics of this dataset"
   • "Find the top 10 products by sales"
   • "Show the monthly trend"
4. The AI analyzes the data and returns results with charts

Troubleshooting

Common Issues

1. Network error - "Failed to fetch"

   • Problem: the frontend cannot reach the backend API
   • Possible causes:
     • The Render service is waking up (cold start)
     • Network timeout
     • CORS configuration issues
   • Solutions:
     • Wait 1–2 minutes for the Render service to fully start
     • Check backend health at https://dataanalyzeagent2.onrender.com/health
     • The frontend will automatically retry three times with increasing intervals
     • Inspect the browser console for detailed errors

2. "Missing optional dependency 'tabulate'" error

   • Solution: ensure requirements.txt includes tabulate==0.9.0; this is fixed in the latest version

3. Render cold start issues

   • Symptom: connection fails on first visit or after long inactivity
   • Solutions:
     • Wait 30–60 seconds for the service to start
     • Hit the backend health endpoint to warm up the service
     • Consider upgrading to a paid Render plan to avoid cold starts

4. Environment variables not applied on Vercel

   • Check:
     • Make sure NEXT_PUBLIC_API_URL is added in the Vercel project settings
     • The value should be https://dataanalyzeagent2.onrender.com
     • Redeploy after changing environment variables

5. CORS errors

   • Symptom: the browser console shows a CORS policy error
   • Solution: the backend is configured to allow the Vercel domain; check domain settings if issues persist

Deployment Checklist

Render Backend Deployment

• Repository connected to Render
• Build command: pip install -r backend/requirements.txt
• Start command: cd backend && python main.py
• Environment variable OPENAI_API_KEY set
• Service type "Web Service"
• Python version 3.8+

Vercel Frontend Deployment

• Repository imported into Vercel
• Root directory set to frontend
• Environment variable NEXT_PUBLIC_API_URL set to backend URL
• Automatic deployments enabled

Testing Steps

1. Backend health check:

   curl https://dataanalyzeagent2.onrender.com/health

   Should return: {"status": "healthy", "temp_dir": "temp_file"}

2. Frontend connection test:

   • Open the frontend application
   • Check the connection status indicator at the top of the page
   • Green means connected, red means failed

3. Functional test:

   • Try sending a simple message without files: "Hello"
   • Upload a CSV file and ask: "Analyze this data"

Environment Variables

Backend

• OPENAI_API_KEY: OpenAI API key (required)

Frontend

• NEXT_PUBLIC_API_URL: backend API URL (required)

License

MIT License

Contributing

Contributions via Issues and Pull Requests are welcome!