Featured Project
Live on both stores · v1.4.0

Jango

Repository namespace · ExpiryMate
An ingredient and household inventory service that tracks expiry dates and recommends AI-generated recipes prioritizing items that expire soon.
Recruiter snapshotRole, decision, and outcome in 30 seconds
6 product screens
Public repository
Launch page
OwnershipProduct · Mobile · Backend · Admin · Infrastructure
Period2026 · App Store and Google Play launch · v1.4.0
OrganizationPersonal Project
01Problem
People missed expiring ingredients because tracking, recommendations, and shared spaces lived in separate flows.
02Key decision
Validated native scanning, billing, and advertising in EAS release builds on both platforms, while aligning mobile, API, and admin contracts, cost limits, and authorization through shared schemas and server-side verification.
03Verified outcome
Released version 1.4.0 on both the App Store and Google Play. Shared refrigerators, bulk photo intake, Jango Plus, and purchase restoration run against live Railway services, with 885 automated tests across 147 files passing.
2 StoresiOS and Android launchedApp Store · Google Play
885Automated tests passed147 test files across Shared, Mobile, and API
LiveAPI/Admin live on Railway
Work you can inspect

Real product screens before explanations

Each caption explains the user problem addressed by the screen.
Project Overview

What I built

Jango is a Korean-first product released on both major stores, organized as a pnpm monorepo containing an Expo mobile app, NestJS REST API, Next.js operations admin, and shared contracts package.The technical packages and bundle IDs retain the legacy ExpiryMate namespace, while the user-facing brand is Jango.
Technologies
React Native
Expo
TypeScript
NestJS
Next.js
PostgreSQL
Prisma
OpenAI
Zod
TanStack Query
Zustand
Docker
Railway
EAS
Sentry
Scope

Key Contributions

  1. 01Built authentication, inventory, recommendation, shared-space, and settings flows with Expo Router, TanStack Query, Zustand, React Hook Form, and Zod.
  2. 02Connected ML Kit barcode and expiry OCR plus OpenAI Vision receipt and refrigerator-photo intake to confidence-aware quick-save and correction flows.
  3. 03Implemented product, inventory, space, recommendation, notification, authentication, privacy, and monetization APIs with owner- and role-based authorization using NestJS, Prisma, and PostgreSQL.
  4. 04Added expiring-first strategies, dietary and allergy validation, semantic repair, model rollout controls, and per-call cost and latency observability to recommendations built on the OpenAI Responses API.
  5. 05Implemented personal, household, and store spaces with owner/manager/member roles and email or one-time-code invitations.
  6. 06Implemented monthly and annual Jango Plus subscriptions, App Store and Google Play purchase verification and restoration, server notifications, and SSV-backed rewards for non-personalized ads.
  7. 07Implemented post-cooking inventory deduction and undo, step timers, and shopping and affiliate flows for missing or depleted ingredients.
  8. 08Built Next.js administration for products, inventory, support, and AI cost and revenue metrics, and configured Railway, Docker, EAS, Sentry, and Resend operations.
Constraints & Decisions

Challenges

  1. 01Native scanning, billing, and advertising unavailable in Expo Go required dev and EAS release builds plus real-device validation on iOS and Android.
  2. 02Shared contracts across mobile, API, and admin had to expose only client-safe types and schemas.
  3. 03Privacy, cost limits, and purchase state for AI recommendations, photo processing, store billing, ads, and invitations had to remain consistent with product behavior and store declarations.
Verified Evidence

Results

01Version 1.4.0 is live on both the App Store and Google Play for iOS and Android.
02The project passes 885 tests across 147 Shared, Mobile, and API test files, with linting, type checking, and environment parity checks maintained as release gates.
03Live API and admin services on Railway support shared spaces, photo AI, Jango Plus billing and restoration, waste-prevention insights, and revenue observability.

Related Projects