🔵 Lumina — Crime Intelligence & Analytical Platform

KSP Datathon 2026 — Karnataka State Police × Hack2Skill Track 2: AI-Driven Crime Analytics & Visualization Platform Deployment: 100% Zoho Catalyst Native

---

Overview

Lumina transforms the Karnataka State Police's fragmented, Excel-based crime records into a real-time, AI-powered Strategic Intelligence Hub — enabling proactive policing, criminal network analysis, and spatiotemporal crime visualization.

Key Features

Feature	Technology
🗺️ State Overview Heatmap	Deck.gl HeatmapLayer
🔴 Hotspot Explorer	ST-DBSCAN on AppSail
🕸️ Criminal Network Graph	Cytoscape.js + Neo4j
📊 Predictive Risk Scoreboard	Zia AutoML + ECharts
💬 AI Query Assistant	Catalyst QuickML (RAG)
📄 One-Click Briefing Reports	Catalyst SmartBrowz

---

Repository Structure

Lumina/
├── catalyst.json               # Catalyst project manifest
├── BackEnd/                    # Backend services, containers, and functions
│   ├── appsail/                # Custom container runtimes
│   │   ├── neo4j/              # Neo4j Docker container + API
│   │   └── ml_pipeline/        # ST-DBSCAN Python runtime
│   ├── client/                 # React + Vite SPA (Web client)
│   └── functions/              # Serverless backend functions
│       ├── api_service/        # Serverless CRUD APIs
│       └── etl_cron/           # Cron-triggered ETL jobs
├── DataBase/                   # Data files, schema, and generation scripts
│   ├── data/                   # Generated dataset files
│   ├── docs/                   # Database schema and docs
│   └── generator/              # Synthetic data generation scripts
├── Documents/                  # Documentation, architecture, and API references
└── FrontEnd/                   # Frontend assets / UI project

---

Quick Start

1. Generate Synthetic Data

cd data/generator
pip install -r requirements.txt
python generate_synthetic_data.py

This produces 7 CSV files in data/synthetic/ covering Districts, Police Stations, FIRs, Accused, Victims, Case-Accused links, and Risk Scores.

2. Run Frontend (Sprint 3)

cd client
npm install
npm run dev

3. Deploy to Catalyst

catalyst deploy

---

Tech Stack

Layer	Catalyst Service
Frontend Hosting	Catalyst Web Client Hosting
Frontend Framework	React 18 + Vite + Deck.gl + Cytoscape.js
API Gateway	Catalyst API Gateway
Authentication	Catalyst Authentication (RBAC)
Backend Compute	Catalyst Serverless (Advanced I/O)
Custom Containers	Catalyst AppSail (Neo4j, ST-DBSCAN)
Relational DB	Catalyst Data Store
Object Storage	Catalyst Stratus
Caching	Catalyst Cache
LLM / RAG	Catalyst QuickML
NLP	Catalyst Zia Services
PDF Reports	Catalyst SmartBrowz
Job Scheduling	Catalyst Cron
Event Routing	Catalyst Signals

---

Documentation

• Full Documentation
• Solution Architecture
• Data Store Schema

---

License

This project was built for the KSP Datathon 2026. All rights reserved.