A production-grade Streamlit data intelligence platform.
# 1. Install dependencies
pip install -r requirements.txt
# 2. Run the app
streamlit run app.pyThen open http://localhost:8501 in your browser.
datavyn/
├── app.py # Entry point
├── style.css # Custom dark neon theme
├── requirements.txt
└── modules/
├── state.py # Session state management
├── sidebar.py # Sidebar with stats & settings
├── overview.py # Landing / overview page
├── upload.py # File upload (CSV/JSON/Excel/Parquet)
├── kaggle_connect.py # Kaggle API integration
├── db_connect.py # SQLite / PostgreSQL / MySQL
├── ai_insights.py # Anthropic Claude AI analysis
├── charts.py # Plotly chart engine
└── export.py # PDF / CSV / Excel / JSON export
| Feature | Description |
|---|---|
| Multi-format Upload | CSV, TSV, JSON, Excel, Parquet |
| Auto Charts | Histograms, scatter, heatmaps, box plots, time series |
| Kaggle Connect | Search & download datasets via Kaggle API |
| Database Connect | SQLite, PostgreSQL, MySQL + Chinook demo DB |
| AI Insights | 5 analysis modes powered by Claude (Anthropic API) |
| Data Cleaning | Drop duplicates, fill NaN, various strategies |
| PDF Export | Styled reports with stats tables + AI insights |
| CSV/JSON/Excel Export | One-click downloads in all formats |
| Service | Where to get |
|---|---|
| Anthropic (Claude) | console.anthropic.com → API Keys |
| Kaggle | kaggle.com → Account → API → Create New Token |
Enter keys via the sidebar (Anthropic) or the Kaggle Connect tab.
- streamlit — UI framework
- pandas / numpy — data processing
- plotly — interactive charts
- reportlab — PDF generation
- kaggle — Kaggle API client
- sqlalchemy — database connectivity
- openpyxl / pyarrow — Excel / Parquet support
DataVyn uses a custom dark neon-industrial theme with:
- Color palette: Deep navy + cyan + violet accents
- Fonts: Syne (headings), JetBrains Mono (data/code), Space Mono (labels)
- 4 switchable chart themes: Neon Dark, Minimal Dark, Vibrant, Monochrome
DataVyn Labs v1.0.0