A real-time stock market surveillance dashboard that detects suspicious insider trading activity across NSE-listed stocks using machine learning, statistical anomaly detection, and regulatory correlation — built for fintech compliance teams.
🔗 Live Demo: https://unmeshasingh.github.io/insiderwatch
⚙️ Backend API: https://insiderwatch.onrender.com
Insider trading costs Indian capital markets billions annually. SEBI receives thousands of complaints every year but manual surveillance is slow, expensive, and reactive. InsiderWatch automates the detection process using a 3-layer AI detection engine that flags suspicious activity before it becomes a scandal.
Layer 1 — Statistical Volume Analysis Calculates 30-day rolling average volume for each stock and flags trades where volume exceeds 3x the average. Red bars on the volume chart indicate suspicious trading windows.
Layer 2 — Pre-Announcement Correlation Cross-references suspicious trades against a corporate events calendar (earnings, mergers, acquisitions, board meetings). Trades occurring 3–30 days before a major announcement are flagged and scored.
Layer 3 — ML Isolation Forest Uses scikit-learn's Isolation Forest algorithm to detect statistical anomalies in trading patterns. Features include volume spike ratio, price change %, volume change %, and price momentum. Model is trained on 3 months of historical data per stock.
Risk Scoring Engine Combines all 3 layers into a 0–100 risk score with full explainability — tells you exactly why a stock was flagged, not just that it was.
- 📊 Live NSE Stock Data — Real-time price and volume charts via yfinance
- 🤖 ML Anomaly Detection — Isolation Forest model trained on historical data
- ⏰ Pre-Announcement Detector — Flags trades before earnings, mergers, acquisitions
- 🕸️ Insider Trade Network Graph — D3.js force-directed graph showing connected trades
⚠️ Risk Score + Explainability — 0-100 score with specific reasons for each flag- 🚨 Live Alerts Feed — Real-time feed of suspicious stocks ranked by risk
- 📋 SEBI Regulatory Reference — Links each flag to SEBI PIT Regulations 2015
- 📄 Compliance Report Generator — One-click downloadable report for flagged stocks
- 🔴 Live Surveillance Ticker — Real-time alert ticker simulating active monitoring
| Layer | Technology |
|---|---|
| Frontend | HTML, CSS, JavaScript |
| Charts | Chart.js |
| Network Graph | D3.js (force-directed) |
| Backend | Python, FastAPI |
| Stock Data | yfinance (NSE/BSE) |
| ML Model | scikit-learn (Isolation Forest) |
| Data Processing | Pandas, NumPy |
| Frontend Deployment | GitHub Pages |
| Backend Deployment | Render |
insiderwatch/
├── frontend/
│ ├── index.html # Main dashboard
│ ├── css/
│ │ └── styles.css # Dark finance theme
│ └── js/
│ ├── dashboard.js # Core dashboard + report generation
│ ├── charts.js # Chart.js price + volume charts
│ ├── network.js # D3.js insider network graph
│ └── api.js # FastAPI integration + real data loader
├── backend/
│ ├── main.py # FastAPI routes
│ └── requirements.txt # Python dependencies
└── README.md
| Method | Endpoint | Description |
|---|---|---|
| GET | / |
Health check |
| GET | /api/stock/{ticker} |
Full analysis for any NSE stock |
| GET | /api/alerts |
Current suspicious stocks ranked by risk |
| GET | /api/watchlist |
List of monitored stocks |
Example response for /api/stock/RELIANCE:
{
"ticker": "RELIANCE",
"current_price": 1293.0,
"risk_score": 25,
"reasons": ["ML model flagged 4 statistical anomalies"],
"suspicious_flags": [...],
"ml_flags": [...],
"prices": [...],
"volumes": [...]
}Frontend:
# Clone the repo
git clone https://github.com/UnmeshaSingh/insiderwatch.git
cd insiderwatch
# Open frontend/index.html in your browser
# Or use Live Server in VS CodeBackend:
cd backend
# Install dependencies
pip install fastapi uvicorn yfinance pandas numpy scikit-learn requests
# Run the server
uvicorn main:app --reload
# API available at http://127.0.0.1:8000
# API docs at http://127.0.0.1:8000/docsRELIANCE | TCS | INFY | HDFCBANK | WIPRO
TATAMOTORS | ADANIENT | BAJFINANCE | ICICIBANK | SBIN
InsiderWatch references the SEBI (Prohibition of Insider Trading) Regulations, 2015 — specifically:
- Regulation 3 — Communication or procurement of unpublished price sensitive information
- Regulation 4(1) — Prohibition on insider trading
- Regulation 9 — Code of conduct for listed companies
This is the same regulatory framework used by SEBI's surveillance department and compliance teams at major financial institutions.
- Expand to 500+ NSE stocks
- Integrate real SEBI insider trading disclosures (Form C filings)
- Add options market surveillance (unusual call/put activity)
- Build email alert system for compliance officers
- Train LSTM model for more accurate anomaly detection
- Add BSE stocks and cross-market correlation
- REST API for integration with existing compliance platforms
This system is designed for:
- Compliance teams at banks and asset managers monitoring suspicious activity
- Fintech startups building regulatory technology (RegTech) products
- Research analysts studying market microstructure and trading patterns
- Regulators looking for automated surveillance tools
Unmesha Singh — CS Student | Kolkata, India
InsiderWatch is built for educational and research purposes. The detection algorithms simulate real surveillance techniques but should not be used as the sole basis for legal or financial decisions. All data is sourced from publicly available market feeds.