Shadow AI Discovery & Risk Intake

Governance-first Shadow AI discovery and triage dashboard built with Streamlit.

Live Demo

🔗 https://shadow-ai-discovery-intake-9rxlhv4tjqxkr3mdwwbfrb.streamlit.app/

Note: Demo uses simulated telemetry and may be temporarily rate-limited during active development.

Overview

This project demonstrates how organizations can:

• Discover Shadow AI usage
• Triage risk using metadata-only signals
• Enable policy-aligned intake rather than reactive blocking
• Generate executive-ready summaries

The demo is privacy-aware by design:

• No prompt or content inspection
• Metadata-first signals only
• Simulated data only

Features

• Risk-level, department, and category filtering
• Executive KPI cards
• Risk distribution and department heatmaps
• Detailed tool-level views
• Exportable CSV and TXT executive summaries

Tech Stack

• Python
• Streamlit
• pandas
• numpy
• plotly

Author

Built by Adarian Dewberry
🔗 Portfolio: https://adariandewberry.com | LinkedIn: https://www.linkedin.com/in/adariandewberry/