NUG Retail

The brief

NUG Retail were going from a single-channel shop to a multi-surface marketplace, and they needed every part of the product to land at the same level of polish: iOS and Android for the bulk of demand, a responsive web experience for desktop and casual shoppers, and an operator dashboard powerful enough that fulfillment, inventory, and delivery dispatch could all be run from one console. The product had to scale from day one because the launch markets already had real volume.

Three buyer surfaces, one product

We treated the iOS, Android, and web buyers as one product team with three deliverables rather than three independent apps. A shared API contract, a shared design system, and a shared analytics schema mean a change in catalog logic ships everywhere in the same release. Native mobile gets the speed and gesture model shoppers expect; web gets the discoverability and shareable URLs that fuel paid acquisition.

The operator console

The admin dashboard is the product's centre of gravity. It's where fulfillment, customer service, inventory, and delivery dispatch all coexist in real time. We designed it as a unified operator workspace rather than a stack of admin pages.

• Inventory and catalog — bulk import, low-stock alerts, supplier reorder triggers, and variant-level pricing rules.
• Order pipeline — a live kanban from new → picked → packed → dispatched → delivered, with SLAs and exception flagging.
• Delivery dispatch — driver assignment, route packing, and live ETA tracking on a single map view.
• Customer service — order-aware chat with one-click refunds, partial returns, and goodwill credits within configurable limits.
• Analytics — cohort retention, AOV by category, fulfillment time distribution, and demand forecasting feeding the inventory queue.

Intelligent dispatch and demand forecasting

Sub-thirty-minute delivery isn't a checkout promise — it's an operations problem. We built a dispatch service that solves a constrained routing problem each time a new order lands: which driver, with which other in-flight orders, in which sequence, given live traffic and prep time at the kitchen. A separate demand-forecasting model projects next-hour and next-day order volume by zone, which feeds driver-shift suggestions and inventory replenishment up the chain.

Discovery and personalization

The buyer apps lean on a recommendation layer that personalizes home-screen rails by purchase history, browsing behavior, and time of day. Reorder prompts surface staples the customer has previously ordered; cross-sell rails feature genuinely complementary inventory the model has learned, not blunt category co-occurrence.

What we shipped

• Native iOS and Android apps with shared design language and full discovery → checkout → tracking flow.
• A responsive web buying experience tuned for SEO and shareable product URLs.
• A unified operator console covering inventory, orders, dispatch, support, and analytics.
• A dispatch and route-optimization service powering sub-thirty-minute delivery at scale.
• A personalization layer driving home feeds, reorder prompts, and cross-sell rails.

The result

NUG Retail launched into a live market and scaled to thousands of daily orders while holding average delivery time under thirty minutes. The operator console replaced what was previously four separate tools; the dispatch engine handles peak hour without an operator manually re-routing; and the apps consistently top their category in store ratings since launch.