See exactly where fraud is happening — down to the district level. vcurd's live heatmap plots every transaction with district-level risk scores, bubble sizing by fraud volume, and automatic MHA high-risk district overlay.
Bubble size = fraud count. Color = risk level. Every transaction scored in real-time adds to the district aggregate.
Toggle between views to answer different investigation questions.
Each district shows a bubble scaled by raw fraud transaction count. Larger bubble = more fraud events. Color encodes the district's risk band (HIGH / MEDIUM / LOW) computed from fraud rate.
All 34 MHA Cyber Crime Portal high-risk districts are pre-loaded and visually distinguished. New fraud in these districts automatically triggers analyst alerts regardless of score.
Switch from count-based to amount-based view. Bubbles resize by total fraudulent ₹ amount per district — surfaces high-value fraud clusters that may have low transaction counts.
Compares last 24h vs. prior 7-day average per district. Rapidly rising districts (velocity spike) are highlighted in orange — early warning before volumes accumulate.
Each district gets a composite risk score based on four signals, updated in a rolling 24-hour window.
| Signal | Weight | Description |
|---|---|---|
| Fraud rate | 40% | Fraudulent transactions ÷ total transactions from this district in last 24h |
| MHA flag | 30% | Binary: district is in MHA Cyber Crime Portal high-risk list |
| Velocity delta | 20% | % increase in fraud count vs. prior 7-day average — detects emerging hotspots |
| Amount concentration | 10% | Average fraud ₹ amount vs. national median — high-value fraud inflates score |
District risk bands: 0–39 = LOW, 40–69 = MEDIUM, 70–100 = HIGH. HIGH districts appear in analyst dashboards and trigger daily summary emails to compliance teams.
Pull district-level fraud aggregates programmatically for your own BI tools, reports, or compliance dashboards.
| Parameter | Type | Description |
|---|---|---|
days | integer | Rolling window in days (1–90). Default: 30 |
state | string | Filter to a single state (e.g. Jharkhand). Optional. |
min_count | integer | Only return districts with at least N fraud events. Default: 1 |
format | string | json (default) or csv |
{
"districts": [
{
"district": "Jamtara",
"state": "Jharkhand",
"lat": 24.18,
"lon": 86.57,
"total_transactions": 412,
"fraud_count": 89,
"fraud_amount": 2340500.00,
"fraud_rate": 0.216,
"risk_band": "HIGH",
"is_mha_flagged": true,
"velocity_delta_pct": 34.2
}
],
"generated_at": "2025-06-01T09:15:00Z",
"window_days": 30
}Daily briefing tool. The morning shift opens the heatmap to see overnight hotspots and allocate analyst capacity to the highest-risk districts before case queues form.
Data-driven rule calibration. If Mewat's fraud rate spikes above 20%, the policy team can activate stricter scoring for that district in the rule engine without a code change.
Generate district-level fraud tables for quarterly regulatory submissions. The CSV export maps directly to RBI's cyber fraud reporting format.
When a new fraud cluster emerges, vcurd's heatmap provides geospatial evidence for the CERT-In incident report — timestamped district data with transaction counts and amounts.
Request a demo with your own transaction data and see your India fraud heatmap in minutes.