AI-powered policy analyzer that finds legal "exploits" in your company's time-off policies — bridge days, holiday stacking, sick leave conversions, and policy loopholes to maximize your days off.
- Upload your company policy document (PDF, DOCX, TXT, or screenshot)
- AI scans the document to extract time-off rules and restrictions
- Exploits detected — bridge days, holiday stacks, policy loopholes with ROI calculations
- Calendar optimizer shows optimal dates to maximize consecutive days off
- Document parsing — PDF, DOCX, TXT, and image OCR via Claude Vision
- Two-tier AI analysis — Free basic scan (Haiku) or Pro deep analysis (Sonnet) for $4
- Country-aware holidays — Integrates public holidays via Nager.Date API
- Exploit detection — Bridge days, holiday stacking, sick conversions, policy loopholes
- ROI calculations — Shows days off gained per PTO day spent
- Calendar view — Visual planner with color-coded days
- Activity history — Track applied/dismissed exploits
| Layer | Technology |
|---|---|
| Framework | Next.js 16 (App Router, Turbopack) |
| Database | Supabase (Postgres + RLS) |
| Auth | Supabase Auth (Email + Google OAuth) |
| Storage | Supabase Storage (policy files) |
| AI | Anthropic Claude (Haiku free / Sonnet pro) |
| Payments | Stripe (one-time $4 per Pro scan) |
| Styling | Tailwind CSS 4 |
| Deployment | Vercel |
# Install dependencies
npm install
# Copy env template and fill in your keys
cp .env.local.example .env.local
# Run Supabase migrations (if using local Supabase)
npx supabase db push
# Start dev server
npm run devOpen http://localhost:3000.
NEXT_PUBLIC_SUPABASE_URL= # Supabase project URL
NEXT_PUBLIC_SUPABASE_ANON_KEY= # Supabase anon/public key
SUPABASE_SERVICE_ROLE_KEY= # Supabase service role key
ANTHROPIC_API_KEY= # Anthropic API key
STRIPE_SECRET_KEY= # Stripe secret key
STRIPE_WEBHOOK_SECRET= # Stripe webhook signing secret
NEXT_PUBLIC_APP_URL= # Your app URL (http://localhost:3000 for dev)
users— Profile with country, PTO balance, plan tierpolicies— Uploaded documents with scan statusexploits— Detected exploits linked to policiescalendar_events— User's planned time-off eventsactivity_log— Audit trail of all actions
All tables use Row-Level Security — users can only access their own data.
Document → Parse (PDF/DOCX/TXT/OCR)
→ Stage 1: Extract policy rules (Haiku)
→ Stage 2: Detect exploits against holidays + PTO (Haiku or Sonnet)
→ Store results
MIT