About
We are building the wallet, the data, and the intelligence in one place.
Feddi is the retail OS for GCC operators. A branded customer wallet that the merchant owns, on top of one consistent intelligence layer that runs across every branch, every till, every campaign.
Why this exists
GCC retail is moving faster than its tooling. Chains that have grown from three branches to thirty in two years are still running customer relationships on point cards, third-party loyalty apps, and the receipt printer. The data the operator needs to actually run the business, who bought what, in which branch, after which campaign, at which margin, sits across four vendors that do not talk to each other.
The result is operators who can describe what they want to do but cannot do it. They want to credit dormant customers with ten dirhams to bring them back. They want to know whether the new Diriyah branch cannibalised the Olaya branch or pulled net new. They want to ask, in plain language, which SKU is selling among repeat buyers, and act on the answer the same day.
That work shouldn't require a data team and a five-vendor stitch. Feddi is what it looks like when the wallet, the rules engine, the attribution layer, and the operator interface are one product built on shared data, by a team that lives in the GCC and ships weekly.
What we believe
-
The wallet is the relationship.
Stored value the merchant owns is the only customer relationship that doesn't depend on a third party's incentive to keep showing it. Points programs and partner apps are leases on the customer. Wallets are ownership.
-
Identification has to be honest.
Loyalty platforms that promise frictionless identification at the till are usually lying about the friction. Feddi requires the customer to be identified the same way a credit card is, explicitly, at the till. The mechanic isn't sleight of hand; it's the operator getting paid upfront and the customer getting something useful in return.
-
Branch-level data or nothing.
"Network-wide repeat rate is 62%" is a marketing slide. "Olaya is at 71% but Tahlia dropped to 48% after the new manager started" is operations. We default to the second, in every dashboard, every report, every brain query.
-
Intelligence should answer the question the operator actually asked.
The decisioning brain doesn't reason about hypotheticals; it reads the data and returns the answer. Approved actions run through the same primitives an operator would have used by hand. The audit trail is identical.
-
Built where it ships.
The team works from the GCC, in the timezones the customers run on, in the languages they think in. We don't ship features tuned for a different market and call them localised after.
How we work
Small team. Operator-first roadmap, where every quarter's plan starts with three customer-shaped problems and works backwards into product. Weekly shipping cadence rather than quarterly releases. Every feature gets used by at least one operator before it lands in the marketing site. No anonymous demo stack, what we show is what we run.
We don't promise vapor. The roadmap items not yet shipped are explicitly marked as in development, with a target window. When the window slips, the page changes the same week.
Team
A small team based across the GCC. Public bios shipping with the next site update.
For partnership, press, or hiring conversations, reach us at hello@feddi.com.
Where we are
The team works across the GCC, with operating presence in Dubai and Riyadh. Sub-processors, data residency, and security posture are detailed at /security/.
For sales: hello@feddi.com. For security: security@feddi.com. For press: press@feddi.com.