Work · uEngage

Four channels. One platform.

Customer experience for more than 200 brands across App, Web, Kiosk and WhatsApp, with the kiosk rewritten from the ground up and the numbers to show for it.

Four channels.
One platform.

uEngage powers ordering for more than 200 brands. Fine dining rooms and quick-service counters, bakeries and pizza chains, cloud kitchens with no front door at all, and campus food courts obliged to move a thousand students between lectures. Each carries its own menu logic, pricing rules, offers and proposition, and each expects the platform to behave as though it were built for them alone.

I joined in May 2026 to own customer experience across the four channels that serve all of them: App, Web, Kiosk and WhatsApp. The mandate is to give every brand the best of what the platform can do, and to ship it at a pace they notice.

The kiosk, rewritten from the ground up

The kiosk came first. The research began at other people's kiosks, at McDonald's, KFC and Burger King, observing what holds up when a hungry person is standing at a screen with a queue forming behind them. I carried the product end to end from there, the feature set and every screen and state, and the team rewrote the codebase from scratch onto a configuration-driven build. The enhanced version shipped inside a month of scoping. The first revamped release went live in mid-August, and three weeks later the entire fleet was running one build, delivered over the air, with no site visits and no interrupted orders.

What shipped: eighteen screens. Six upselling screens, each configurable by brand, where the previous kiosk had none. Offers brought to the front, with one-tap coupons and combination builders. No empty screens: a search returning nothing presents curated selections rather than a dead end. A composed, specific screen for every payment outcome, so that a failed payment becomes a recovery rather than an abandoned order. And customer mobile capture, the foundation everything loyalty needs.

Built for 200 brands

A pizza chain wants combinations. A bakery wants its bestsellers surfaced first. A cloud kitchen wants the shortest possible path to payment. A campus wants a queue to keep moving. The revamp treats all of that as configuration rather than code. A brand's menu, offers, propositions, payment methods and theme are set from a management dashboard, and the same build runs everywhere. Onboarding a brand or reconfiguring an existing one requires no app release, which is how the platform manages to feel bespoke to each of them and still ship weekly.

Measured like a product

Before the revamp the kiosk had never reported a funnel at all. It now reports 79 events across a nine-step journey, with brand and outlet identity carried on every one, and a weekly report, written once, that runs itself every Monday: scorecard, funnel, payment economics, reliability, and a flag engine that converts thresholds into a work queue. Test traffic is excluded at source, so every figure is defensible by construction.

The cleanest reading is like-for-like: the same kiosks, the weekend after launch set against the weekend before. Conversion up 4.9 points. Average order value up 25 percent. Revenue per session up 36 percent. Payment success up 3 points. Over the following four weeks weekly sales grew 21 percent on the same fleet, and between May and August kiosk GMV roughly doubled while kiosk orders grew about 60 percent.

The figure I watch most closely is the one that points forward. From first add to paid order, the kiosk converts at 84 percent, and it has held there since launch. Intent converts. The entire opportunity lies in getting the first item into the cart, which is precisely what the upselling screens and a merchandised home screen exist to do.

App and Web, argued against numbers

The app received a KPI tree first, so that every feature argument could be settled against a number rather than an opinion. Then a master specification for every post-order screen at mobile and desktop, feature toggles including a universal clean-mode filter, and a feature guide so the sales team can see at a glance what the platform offers, what a brand can configure for itself, and what genuinely requires development.

WhatsApp, read from the code

I mapped every journey of the WhatsApp ordering storefront from the production code, benchmarked it against Meta's platform capabilities, and wrote the build specification for webview ordering with a companion document on abuse prevention.

Loyalty and upselling

A national burger brand needed a complete earn-and-burn loyalty experience across app and web, for first-time and returning members alike. I designed it and handed it to engineering. Across the platform, the same thinking became a cross-channel upselling engine, a single service recommending across App, Web and Kiosk, now documented and in review. The kiosk data already makes the case twice over: orders influenced by a recommendation are worth about 46 percent more than menu-only orders, and once a first item is in the cart, 84 percent of sessions pay.

What the platform did

Between May and August, platform GMV grew about 13 percent and orders about 10 percent, with dine-in the fastest-growing order type at about 31 percent. The kiosk roughly doubled. I would not claim all of that as mine. I would claim the measurement that allows us to say it with confidence.

What it taught me

Clarity ships faster than headcount. Decisions in writing, one build across the fleet, and a number behind every claim.

Hi, I'm Parth's AI.
Ask me anything, or I'll show you around.
Tour · 1 of 10
space · ← → · esc

That's the whole story.

If any of it is useful to you, the next step is a conversation.

Book a call