Data Analysis Agent — Full Stack

A web application for interactive data analysis powered by a PydanticAI agent with real-time streaming.

Features

• Natural language queries over CSV datasets via DuckDB SQL
• Real-time streaming of the agent's reasoning (thinking blocks), tool calls, and responses
• Interactive Plotly visualizations rendered directly in the chat
• Data tables for tabular results
• Tool call transparency — see the agent's tool calls, arguments, and results as they happen

Demo

---

Stack

• Backend: FastAPI + PydanticAI + DuckDB, streamed via SSE
• Frontend: React + TypeScript + Vite + Tailwind CSS
• Infrastructure: Docker Compose

Getting Started

Prerequisites

• Docker & Docker Compose
• An Anthropic API key

Launch

# 1. Set your API key
cp .env.example .env
# Edit .env and add your ANTHROPIC_API_KEY

# 2. Add CSV files to data/

# 3. Start the web app
docker compose --profile web up

The app will be available at http://localhost:3000.

The data/ directory is mounted as a volume — you can add or modify CSV files without rebuilding.