Nextgrab

The problem

The first version had been put together quickly with AI. It looked like an app, but it didn’t hold up once real customers and vendors started using it.

From shelf to pickup

How one surprise bag moves through the marketplace, from a business with food left over to a customer walking out with it.

1. 01 · Vendor app & dashboard

   A business lists what’s left

   Vendors add surprise bags with a pickup window, from the vendor app or the web dashboard, with staff accounts across several stores.

   Multi-store staff · vendor tiers
2. 02 · Customer app

   Customers find deals nearby

   The map loads stores by zoom level and distance, and lists are paginated, so it stays fast on real phones. People browse as guests and only sign in at checkout.

   10 km default radius · Google & Apple sign-in
3. 03 · Checkout

   Reserve and pay in seconds

   Pickup or delivery, paid by card, Whish or cash on pickup.

   Card via Stripe · Whish · cash
4. 04 · Pickup

   A QR code confirms the handover

   The customer shows a code and the vendor scans it in the app or the dashboard. The customer’s impact stats update, and admin and vendors share one record of who owes whom.

   Meals rescued · money saved · CO₂ prevented

Key decisions

1. 01

   People can browse as guests and only sign in when they check out, with Google and Apple sign-in, password reset by OTP and account deletion built in.

2. 02

   Each customer’s profile tracks their impact (meals rescued, money saved and CO₂ prevented), and the new app ships fixes over the air with code push, with Crashlytics reporting what breaks in the wild.

Outcome

Turned a prototype into a marketplace real people use: I rebuilt the customer app (shipped as version 2 on the App Store) and the website, and built the vendor app, vendor dashboard and admin. 5K+ downloads on Google Play.