Political prediction marketplace targeting Kenyan 2027 elections. Politician-only platform using M-Pesa payment integration and Kalshi-style YES/NO prediction contracts.
- Platform Type: Binary prediction market (YES/NO contracts)
- User Base: Politicians only (pay-to-play access via M-Pesa)
- Monetization: Entry fees (KES 50,000) + boost credits + prediction pools
- Timeline: 18-month campaign season cash machine (Early 2026 - Dec 2027)
- Data: Simulated/anonymized (no real user scraping liability)
- Backend: Node.js + TypeScript + Express + Socket.IO
- Database: PostgreSQL + Redis
- Frontend: React + Vite + Tailwind CSS
- Payment: M-Pesa API integration + webhook verification
- Deployment: Docker + Docker Compose
- Entry Fee: KES 50,000 for politician platform access
- Boost Credits: KES 1K/5K/10K packages (24-72hr duration, auto-decay)
- Prediction Pools: 15% house edge on all YES/NO contracts
- Rival Data Unlocks: KES 2,000 per competitor analysis
- Anti-Decay Protection: KES 50,000/month subscription
- Decay Mechanics: Boosts expire constantly → repeat purchases
- Rival Alerts: Realtime notifications when competitors spend
- Leaderboard Pressure: Visual ranking triggers status anxiety
- Campaign Sensitivity: Auto-adjusts decay/alerts/fees as elections approach
- Addiction Loops: Dashboard designed for compulsive checking
- Environment variables only (never commit secrets)
- All payments via M-Pesa webhooks with transaction logging
- Immutable audit logs for regulatory questions
- Politician KYC verification required before platform access
- Rate limiting + fraud detection on payment flows
# Development setup
docker-compose up -d # Start DB + Redis
cd apps/backend && npm run dev # Backend on :3001
cd apps/frontend && npm run dev # Frontend on :3000
# Database
npm run migrate # Run schema migrations
npm run seed:politicians # Create test politician data
npm run seed:markets # Generate prediction markets
# Testing
npm run test # Unit tests
npm run test:e2e # End-to-end payment simulation