MEZA — Experience Intelligence Platform

Restaurant experience intelligence: understand how occupancy, environment, and operational conditions influence customer behavior and revenue, using anonymous occupancy analytics and revenue data. Not a POS system — MEZA reads signals (CCTV-derived occupancy, environment, POS exports) and turns them into dashboards, experiments, and recommendations.

Try it in 2 minutes

No signup, no camera, no setup:

1. Open the deployed app (or run npm install && npm run dev locally).
2. On the sign-in page, click Try Demo.
3. You're in a read-only demo restaurant ("Meza Demo Bistro") with 60 days of realistic occupancy, environment, and revenue data already loaded.
4. Click through Dashboard → Occupancy → Environment → Experiments → Recommendations — every chart and table is populated, including a completed experiment with real measured results.
5. Nothing you click can modify the data (enforced by Row Level Security, not just the UI) — explore freely.

If the demo button isn't visible, it hasn't been configured for this deployment yet — see Demo mode below to set it up.

Screenshots

Screenshots aren't checked into this repo yet — this session's sandbox couldn't capture them (headless preview rendering timed out). To add them: run npm run dev, sign in via Try Demo, and capture /dashboard, /occupancy, /environment, /experiments, and /recommendations into docs/screenshots/, then reference them here, e.g.:

![Dashboard](docs/screenshots/dashboard.png)
![Occupancy Analytics](docs/screenshots/occupancy.png)

Features

• Occupancy Analytics - Real-time table and people counting from existing CCTV
• Environmental Tracking - Temperature, music, lighting, weather correlations
• Revenue Analytics - Daily revenue trends and order analysis
• Experimentation - Design and track A/B tests for environment optimization
• Recommendations - Data-driven suggestions for improving experience

Privacy

This system does NOT collect:

• Customer names, phone numbers, or IDs
• Facial recognition or biometric data
• Images (processed and discarded immediately)
• WiFi device tracking

This system DOES collect:

• Anonymous occupancy counts
• Table utilization metrics
• Queue length and wait times
• Environmental conditions (temperature, music, lighting)

Setup

1. Create a Supabase project at https://supabase.com
2. Run the SQL migrations in order: supabase/migrations/001_initial_schema.sql, then 002_cameras.sql, then 003_demo_mode.sql (or paste the combined supabase/setup.sql)
3. Copy .env.example to .env.local and add your Supabase credentials
4. Install dependencies: npm install
5. Run development server: npm run dev

Demo mode

MEZA can seed a fully-populated, read-only demo restaurant so anyone can explore the product without signing up or connecting real data:

# .env.local also needs SUPABASE_SERVICE_ROLE_KEY (Project Settings -> API)
npm run seed:demo

This creates one demo auth account and one restaurant flagged is_demo = true, then generates 60 days of occupancy, environment, revenue, experiment, and recommendation data shaped like a real restaurant (weekday/weekend patterns, lunch/dinner peaks). Set NEXT_PUBLIC_DEMO_EMAIL / NEXT_PUBLIC_DEMO_PASSWORD (same values used above) to make the Try Demo button appear on /signin. Read-only is enforced by 003_demo_mode.sql's RLS policies at the database layer, not just in the UI — see docs/DEPLOYMENT.md for the full walkthrough.

Deployment

One-click-ish deploys to Vercel (vercel.json) or Render (render.yaml Blueprint) are configured in-repo. Full step-by-step guide, including Supabase setup and post-deploy verification: docs/DEPLOYMENT.md.

Architecture

• Frontend: Next.js 14 (App Router) + React + TypeScript
• Database: Supabase PostgreSQL with Row Level Security
• Auth: Supabase Auth (email/password)
• CV Pipeline (optional hardware add-on): Python script for edge device (Raspberry Pi / Jetson Nano) — has a --simulate mode that needs no camera; see cv_pipeline/README.md
• Charts: Recharts

Project Structure

app/
  api/           # API routes (restaurants, occupancy, revenue, etc.)
  signin/ signup/ create-restaurant/  # Auth & onboarding pages
  dashboard/     # Main dashboard
  occupancy/     # Occupancy analytics
  environment/   # Environmental tracking
  experiments/   # Experiment management
  recommendations/ # Data-driven recommendations
  cameras/       # CCTV camera & table-region configuration
  upload/        # POS CSV import
components/      # React components
lib/            # Supabase client, types, store, API client
cv_pipeline/    # Python CV pipeline for edge devices (optional add-on)
scripts/        # Demo data seed script
supabase/       # Database migrations

MVP Deployment

• Cost: Free to launch (Vercel/Render + Supabase free tiers). Optional CV hardware add-on: under ₹10,000 (Raspberry Pi 4 + cables)
• Timeline: Live in minutes with demo mode; 2-4 weeks for a camera-equipped pilot
• Data: Manual entry and POS CSV import work with zero hardware; existing CCTV cameras are an optional automation layer
• Privacy: Zero PII collected