Case study · Fintech

Which card to use, at the counter you are standing at

Most people in the UAE hold several cards and use whichever one is nearest the front of the wallet. The rewards they are entitled to go unclaimed, not through indifference, but because nobody can hold twenty-seven banks' reward tables in their head at a till.

27+ banks covered across the UAE Geolocation at the merchant AI recommendations

A decision made at the wrong moment

The choice of card is made in about two seconds, at a counter, with no information to hand. Any product that solves this has to answer before the moment passes, which rules out anything the user has to open and search.

What we built

A geolocation engine that works out which merchant you are near and recommends the card that pays best there, with AI recommendations across 27+ UAE banks. Automated reminders handle the things a cardholder forgets, and the whole thing integrates with the client's CRM so the platform is a channel rather than an island.

Where the AI runs, plainly

This platform uses Gemini and GPT-4o, and location data comes from the Google Places API. It is not self-hosted, and we would rather say that here than let the site's general position imply otherwise. Recommendation quality mattered more on this build than hosting did, and that was the right call for this product.

Stack

React and TypeScript on the front, Node and Express behind it. Gemini and GPT-4o for the recommendations, Google Places API for merchant resolution.

Want the same thing looked at in your business?

We map the process, find the part worth automating first, and give you a written scope with the phases separated. No obligation, and you keep the plan.