Made By You: Every painted piece tracked from drop-off to pickup

01 / Overview

About Made By You

02 / The challenge

What Made By You needed to solve

Between drop-off and pickup, pieces look alike on a shelf, customers forget, and staff were texting people by hand when a piece was ready.

1. 01

   Fired pieces look alike, so every piece needed its own order tied to the customer who painted it.

2. 02

   The studio wanted a second opinion that the fired piece is the one that was ordered before anyone is told to collect it.

3. 03

   Pickup messages had to go out in words the studio controls, not typed one by one by staff.

4. 04

   Pieces left too long had to surface on their own, and counter staff, managers and owners each needed different controls.

03 / What we built

How we built Made By You

We built three TypeScript projects: an Expo app for staff and customers, a Next.js admin console and an Express API on MongoDB. Staff photograph each piece before and after firing, and a person approves every job before the customer hears about it.

• 01

  An order for every piece

  Each piece gets an order with the customer, item, description and table number, and moves through statuses such as in progress, firing, notified, picked up and overdue. Staff search orders by name, email or item and filter them by status and date.

• 02

  Before and after photos

  Staff capture each piece in the app before and after firing. Photos go straight from the phone to DigitalOcean Spaces through short-lived signed links.

• 03

  AI match check

  The API sends both photos and the order description to GPT-4o vision, which returns a confidence score out of 100, a match verdict and a list of concerns, such as a glaze color slightly darker than the before photo.

• 04

  Human review queue

  Every bisque job waits for a person to approve, reject or hold it, in the web console or on the phone. The AI call gets one retry, and if it still fails the job is saved with an error flag for a person to judge.

• 05

  Pickup messages in the studio's words

  Approval fills in the studio's message template with the customer's name and the item, moves the order to notified and sends a text through Twilio and a push message through Expo. A notifications log shows every message with its channel and delivery status.

• 06

  Customer view

  Customers see their approved before and after photos and a ready for pickup notice, and can tap I'm on my way to tell the studio they are coming.

• 07

  Daily overdue sweep

  A daily job at 9 am marks notified orders as overdue after seven days, so forgotten pieces show up without anyone checking the shelf.

• 08

  Roles and studio settings

  Four roles and nine permissions decide what each team member can do. Settings hold the overdue period, the AI confidence cutoffs, the default pickup message and switches for texts and push messages, and deactivated customers are signed out and blocked from the app.

04 / The results

What changed for Made By You

The platform now covers a piece from drop-off to pickup: the order, two photos, an AI check, a human approval and a message to the customer.

• Every piece has an order and two photos, so staff can tell whose piece is whose.
• The AI check gives a scored second opinion on each fired piece, and staff keep the final say.
• Customers hear that their piece is ready by text and push message, in words the studio controls.
• A daily sweep flags pieces left behind, so they don't sit on the shelf unnoticed.
• Customers see their approved photos and can tap to say they are on the way.