Skip to content
KazuhiroxvPublic

About

AI-assisted academic mobile café ordering and management app built with Expo, React Native, TypeScript, and Supabase.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Repository files navigation

Cafinity

Cafinity is an academic multi-role café ordering and management application built with Expo, React Native, TypeScript and Supabase/PostgreSQL. It covers customer ordering, staff/barista fulfilment and administrator management, with access controls enforced in the database, server-side order functions and realtime updates.

Development was substantially AI-assisted. Most of the application code, SQL, documentation and visual assets were produced with Claude Code, under the direction and review of the project lead. See AI-assisted development.

This is a preserved academic version. It is not deployed to production, not published on any app store, has no real users, and processes no real payments.

Academic context

  • Course: CSS152P, Software Engineering 2, Mapúa University (Group 3)
  • Period: June 2026

My role

Project Lead / Development Coordinator. I directed requirements, feature scope, application workflows, iterative review and integration, and used Claude Code extensively for implementation support. I did not hand-write the application source.

What the project implements

Implemented in the source and database. See Verification status for what has been run.

Area Implementation
Customer Sign-up and sign-in, branch selection, menu browsing, drink customization with live pricing, cart, checkout with promo codes, order tracking, order history, receipts, favorites, loyalty points, rewards and vouchers, notifications, feedback
Staff / barista Order queue for the barista's assigned branch, order status updates (pending → preparing → ready → completed), cash-payment confirmation, estimated pickup times
Administrator Dashboard, sales and feedback reports with export, menu and product management, per-branch inventory, user role management, seasonal campaigns, cancellations, app settings. Promo codes are seeded in the database; there is no in-app promo editor.
Backend 29 PostgreSQL tables, all with row-level security; role checks in database functions; prices computed on the server during checkout; realtime subscriptions on orders and notifications

Payments: GCash is a simulated payment path for demonstration purposes. There is no payment gateway and no real money is transferred. Cash orders are confirmed by staff.

Architecture

Expo / React Native app (TypeScript, Expo Router, Zustand, NativeWind)
        │  supabase-js  (publishable key only)
        ▼
Supabase
 ├── Auth          sign-up / sign-in
 ├── PostgREST     tables + RPC functions (place_order, advance_order_status, report_orders, …)
 ├── Realtime      live order status and notifications
 ├── Storage       product images, profile avatars
 └── PostgreSQL    29 tables, row-level security, role-checking functions

Details: docs/architecture.md and docs/database.md.

Tech stack

Expo SDK 54 · React Native 0.81 · React 19 · TypeScript · Expo Router · Zustand · NativeWind · Supabase (Auth, PostgREST, Realtime, Storage) · PostgreSQL 17

Local setup

Requirements: Node.js 20.19 or later, the Expo Go app (SDK 54) on a phone, and either a Supabase project or a local Supabase stack (Supabase CLI and Docker).

  1. Install dependencies:

    npm install
  2. Create the database. Apply the SQL files in supabase/ in the order listed in docs/database.md. Each file must be run in order. Then run supabase/verify_migrations.sql, a read-only check where every row should report ok = true.

  3. Configure the app:

    cp .env.example .env

    Set EXPO_PUBLIC_SUPABASE_URL and EXPO_PUBLIC_SUPABASE_ANON_KEY to your project's URL and publishable (anon) key. Never put a service-role or secret key in the app.

  4. Start the app and open it in Expo Go:

    npx expo start
  5. Create accounts through the app's sign-up screen. Everyone signs up as a customer. Promote staff and admin accounts in SQL:

    update public.users set role = 'admin' where email = 'admin@cafinity.test';
    update public.users
       set role = 'staff',
           branch_id = (select id from public.branches where name = 'Cafinity Intramuros')
     where email = 'barista@cafinity.test';

No demo accounts or passwords are included in this repository.

Verification status

What has been checked for this portfolio version, and what has not:

Check Result
TypeScript (tsc --noEmit) Passes, 0 errors
ESLint 0 errors, 0 warnings
Database from scratch In an isolated local Supabase environment (PostgreSQL 17.6), all 37 SQL files applied unchanged, in order, with no errors
Migration checks supabase/verify_migrations.sql: 17 of 17 checks passed
Live database 29 tables (row-level security on all 29), 44 policies, 55 functions
End-to-end runtime Not yet verified. The customer → barista → administrator order flow exists in code but has not been demonstrated end to end.

The local verification confirms the database setup, not the complete mobile workflow.

Known limitations

  • Web build: the app targets Expo/React Native. During portfolio verification, the configured Expo web static output (web.output: "static") failed during Node pre-rendering. Client-side AsyncStorage access depends on window, which does not exist in Node. This web-specific issue remains unresolved in this preserved version. The mobile (Expo Go) path does not use pre-rendering.
  • Push notifications are local notifications; remote push is not available in Expo Go.
  • Android adaptive icon and favicon still use Expo's default template artwork.
  • EAS builds would need linking to your own Expo account (eas init).
  • The in-app "Contact support" link uses the placeholder address support@cafinity.app. It is not a monitored mailbox, and the project does not own that domain.
  • Three Expo packages are one patch version behind the SDK's recommended versions (for example expo 54.0.35 vs ~54.0.37).
  • No automated test suite.

AI-assisted development

Summary. Full details are in docs/ai-assisted-development.md.

  • Directed by me: requirements, feature scope and scope changes, the customer / barista / administrator workflows, iterative review, integration, build configuration and demo preparation.
  • Produced with Claude Code: most of the application code, the SQL schema and functions, the original project documentation, and most of the visual assets. The images were generated with OpenAI image generation, and 88 of them carry embedded C2PA provenance metadata.
  • Portfolio verification: TypeScript and ESLint checks, and recreating the database from scratch in a local Supabase environment. Both were done after the course for this public version.

Assets

  • Every image except the Expo template icons is AI-generated: products, campaigns, branding, the logo, empty states, the default avatar and the notification icon. See docs/ai-assisted-development.md.
  • The Android adaptive icon and favicon are Expo template defaults.
  • One product image with unknown provenance was removed from this public version. That product falls back to the app's placeholder tile.

License

No open-source license has been granted. This is an academic team project, and no license will be added without the agreement of the team.

About

AI-assisted academic mobile café ordering and management app built with Expo, React Native, TypeScript, and Supabase.

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages