A Gamified Approach to Property Safety | Built by The Mangos
FloorSense application looks at floor plans the way a security expert would, borrowing the same spatial thinking used in tactical games like Rainbow Six Siege and CS:GO to assess real architectural vulnerabilities. How many entry points? Blind corners? Sightlines from the street? Paste a Rightmove link, and our model analyse the floor plan automatically, scoring the property with a Defense Grade. You also get an audio briefing, voiced by ElevenLabs that walks you through exactly what makes the layout strong or exposed.
- Link-to-Intel Pipeline: Drop a Rightmove URL, and we handle the scraping, image processing, and AI analysis automatically.
- Gamified Safety Metrics: We calculate Choke Points, Flank Vulnerabilities, Sightline Dominance, and Defensible Positions.
- Tactical Audio Briefings: ElevenLabs reads out your customized property threat assessment in real-time.
- Interactive Dashboard: A sleek, responsive React + TypeScript frontend to visualize your stats and floor plans.
- Frontend: React, TypeScript (Fast, type-safe web UI).
- Backend: Node.js, Kotlin.
- AI Brain: Anthropic Claude (Vision analysis of floor plan layouts).
- Voice Engine: ElevenLabs API (Text-to-speech mission briefings).
- Database: MongoDB Atlas (Storing property dossiers and tactical stats).
- Infrastructure: Vultr & Docker (Containerized for seamless, scalable deployment).
Here is the step-by-step setup for our Node.js and React environments.
First, you'll need a MongoDB cluster. Create a .env file in both your backend and frontend directories.
Backend .env:
PORT=80
MONGODB_URI=mongodb+srv://<username>:<password>@cluster.mongodb.net/floorsense
ANTHROPIC_API_KEY=your_claude_key
ELEVENLABS_API_KEY=your_elevenlabs_key
Team Members:
Adarash Ram Sivakumar,
Adit Sharma,
Sai Raghav Commandur,
Ervin Oks