# One platform, built from zero Qubriux · Skellam – case study by Seemant Shekhar. https://www.seemant.design/work/one-platform-built-from-zero Beyond onboarding, I designed all of Qubriux: navigation, segments and every builder. Each builder starts with the merchant’s first question: campaigns with *who*, offers with *what*, forms with *when*. - **Role**: Sole product designer - **Team**: CEO, 1 PM, 3–4 engineers - **Timeline**: 2021–2022 - **Scope**: Web app + 2 mobile apps - **Areas of influence**: Navigation and segments; Campaign, offer and form builders; AI Autopilot for games; Co-marketing - **Highlights**: Targeting Tactics, singled out by the client; Campaign 3.0: menus open where you click; AI that merchants can switch off ## 01. Segments with a plan **Every preset segment came with Targeting Tactics.** Active, At Risk, Loyal, Big Spenders: each came with the intent for that segment and three concrete moves. The client singled this out. - Image: Targeting Tactics dialog for a segment with three suggested tactics. Targeting Tactics: the intent, then three moves. ## 02. Campaign 3.0 **The choice appears where you click.** My last redesign before leaving. The + opens its menu right on the branch instead of a panel across the screen, so every step of a long flow is a shorter trip. ### The principles behind it - **Proximity**: The menu sits next to the step it changes. - **Progressive disclosure**: Options appear only when you ask for them. - **Fitts's law**: A shorter trip from the click to the choice. - Image: Campaign 3.0 flow with an If/Else condition and a menu opened inline at the plus button. Designed and handed over at the end of my time there. ## 03. AI on autopilot **A shopping game the AI can run for you.** A timed shopping game for stores. Autopilot lets AI pick steps, products and timings per customer, pushing best-sellers or slow-moving stock. Merchants can switch it off and set everything by hand. - Image: Game configuration with the Autopilot toggle on and the settings AI chooses. AI does the work; the merchant keeps the switch. ## 04. Co-marketing **Stores promoting each other.** Spend ₹1,500 at one store, get 10% off at its partner. The first screen explains it in one sentence and offers two paths: invite a store, or add one already on Qubriux. - Image: Co-Marketing empty state with two options: invite a store partner, or add an existing one. One idea, two clear paths. ## 05. Qffer **The same offers, at the store counter.** Two mobile apps for physical stores: merchants run offers, shoppers find them nearby and redeem them. Read the Qffer case study → - Image: Qffer shopper home with nearby stores and a Scan QR button. Shopper home - Image: Qffer cart with an offer applied. Offer in the cart ## 06. What I’d change **Two lessons, both about keeping context visible.** The form preview was on the wrong side.** I put the preview left and the controls right. Merchants lived in Shopify’s theme editor, with settings on the left, so it broke Jakob’s law. Targeting Tactics should have been a popover.** The modal covered the segment rules the merchant was trying to understand. Preview Controls What I shipped Controls Preview What I’d ship now --- Next case study: [Skip the line](https://www.seemant.design/work/skip-the-line)