Drishti

AI Event Safety Platform · Intelligent event safety with live crowd monitoring, anomaly detection, and emergency dispatch — scalable for 5M+ attendees. AI heatmaps + Gemini analytics cut response time by 35%.

WORLD

Watchtower of Inferno Vale

Live thermal feeds and AI alert beacons sweeping a million-strong valley.

TILESET

Crowd Monitor

Live density tracking

Anomaly AI

TensorFlow + OpenCV

Emergency Dispatch

Instant routing

Heatmaps

Real-time pressure view

Gemini Analytics

GenAI insights layer

WebSocket Bus

Low-latency sync

Alert Engine

Tiered notifications

Replay Vault

Post-event review

ASSETS

Ops Console

Unified command UI

Field Mobile App

Ground team checks

Camera Grid

Multi-feed intake

Risk Score

AI threat scoring

Incident Log

Audit timeline

Capacity Sim

5M+ attendee model

MISSION BRIEF

PROBLEM: Large-scale events lack a unified, real-time view of crowd risk and dispatch needs.

SOLUTION: An AI-driven safety command center that fuses computer vision, crowd analytics, and live ops messaging.

OUTCOME: 5M+ attendee scalability with 35% faster emergency response and richer post-event insights.

KEY HIGHLIGHTS

Built an intelligent event safety platform with crowd monitoring, anomaly detection, and emergency dispatch — scalable for 5M+ attendees
Integrated OpenAI / Gemini analytics with TensorFlow + OpenCV pipelines for live heatmaps and safety alerts
Architected MERN backbone with WebSockets for real-time field sync and low-latency dispatch
Cut emergency response time by 35% through AI-driven prioritization of alerts

PROJECT STATS

5M+

SCALE

-35%

RESPONSE

Live AI

DETECTION

TECH STACK

React.jsNode.jsMongoDBWebSocketsOpenCVTensorFlowGenAI

▶ VIEW CODE