Shopping optimizer
RepMax
Splits a Represent basket across orders to squeeze the most out of Prestige loyalty discounts.
- Client
- Side project
- Type
- Shopping optimizer
- Year
- 2026
- Status
- In use
The brief
Where it started.
Represent's loyalty program allows one discount per order, rewards have minimum-spend gates, and sale items never discount. Splitting a basket into several orders can stack more savings — but working it out by hand is tedious.
What we built
So we shipped 4 pieces.
- 01Exact optimizerSearches every way to split the basket, in a Web Worker
- 02Two modesSmart mode picks rewards from your points; manual mode locks specific codes
- 03Live catalogueSynced from 11 regional storefronts on a schedule
- 04Everyday polishCurrency conversion, shareable scenarios and a guided phone setup
Under the hood
The part worth a closer look.
The clever part
For typical baskets the recommended split is provably optimal, not a heuristic — it scores every partition against tier and reward rules.
Stack
And what it all runs on.
- Next.js 15
- React 19
- Web Workers
- GitHub Actions
- Vitest
Need something like this?
Tell Echo what you’re trying to build — rough is fine. You’ll leave with a clearer sense of what it would take.
← PreviousProof2PrintDesktop app
Next →Japan Trip PlannerPhone-first PWA