Data-Analyst-Agent-API

An AI-powered API that can source, prepare, analyze, and visualize data based on plain English instructions.
This project automates end-to-end data analysis — from fetching datasets to returning ready-to-use insights and visualizations.

---

Key Features

✅ Accepts natural language queries for analysis
✅ Handles multiple input formats (.csv, .json, .png, .jpg, .txt)
✅ Scrapes & fetches data from external sources if required
✅ Performs data cleaning, transformation, and statistical analysis
✅ Generates charts & visualizations (Base64-encoded images)
✅ Returns results in the requested format (JSON arrays, objects, or data URIs)
✅ Deployed on Render for reliable cloud hosting

---

⚙️ Tech Stack

• Backend: Python + FastAPI
• LLM Integration: OpenAI / Groq / OpenRouter
• Data Processing: Pandas, NumPy, DuckDB
• Visualization: Matplotlib, Seaborn
• Deployment: Render Cloud

---

Example Request

curl "http://localhost:8000/api/" \
  -F "questions.txt=@questions.txt" \
  -F "data.csv=@dataset.csv"

---

Example Use Cases

💰 Movies Data – Scrape Wikipedia for highest-grossing films and analyze trends
⚖️ Judgement Data – Query and visualize 1TB+ of Indian High Court judgments
📈 Sales & Finance – Correlation analysis, forecasting, regression plots
🌦 Weather – Historical trend visualization, anomaly detection

---

Run Locally

1️. Clone the repository

git clone https://github.com/ckritk/Data-Analyst-Agent-API.git
cd data-analyst-agent

2️. Set up environment variables
Copy the ENV_TEMPLATE file to .env and fill in your keys/config:

cp ENV_TEMPLATE .env

3️. Install dependencies

pip install -r requirements.txt

4️. Run the application

bash start.sh

The app will now be running locally at:

http://127.0.0.1:8000

---

License

This project is licensed under the MIT License.