Chat with your data in plain English. Upload CSV, Excel, JSON, or SQLite — ask questions and get answers with auto-generated charts.
- Upload your data (CSV, Excel, JSON, SQLite)
- Ask questions in plain English or Arabic
- AI writes pandas code — Groq generates, executes, and self-heals
- Get results — tables, statistics, charts, and insights
- Multi-source: CSV, Excel (.xlsx), JSON, SQLite, PostgreSQL
- Natural language queries: "Show top 5 rows", "Average salary by department", "Plot revenue trend"
- Code generation: AI writes pandas/matplotlib/seaborn code
- Self-healing: If code fails, AI fixes it automatically
- Charts: Bar, line, pie, scatter, histograms — auto-generated
- Conversation history: Remembers context across questions
- Export: Download results as CSV, charts as PNG
- Dashboard mode: "Create a sales dashboard" — generates multi-chart view
- CLI + Web UI: Both interfaces fully functional
pip install -r requirements.txt
# Web UI
python webui.py
# Opens at http://127.0.0.1:8080
# CLI
python cli.py load --file data.csv
python cli.py ask 1 "Show me summary statistics"
python cli.py chat 1 # Interactive modeGet a free Groq API key at https://console.groq.com
set GROQ_API_KEY=gsk_your_key_here| You ask | AI does |
|---|---|
| "Show top 10 rows" | df.head(10) |
| "Average salary by department" | df.groupby('dept')['salary'].mean() |
| "Plot monthly revenue trend" | Line chart with matplotlib |
| "Find correlations" | Heatmap with seaborn |
| "Filter where sales > 1000" | df[df['sales'] > 1000] |
- Backend: FastAPI + Python
- AI: Groq Llama 3.3 70B
- Data: Pandas, Matplotlib, Seaborn
- Frontend: Server-rendered HTML + Chart.js
MIT