Uride
Rider and Driver apps for a Canadian intercity ride-share network where drivers work in shifts. Designed from zero to replace a web service; 1M+ rides across Ontario and BC.
- Role
- Product Designer, later the sole designer
- Team
- Product owners, an ex-Uber project manager and company leadership
- Platforms
- iOS and Android, 320+ screens across two apps
Uride focuses on trips between cities. Unlike typical gig platforms, its drivers are part of the fleet and work scheduled shifts. Backed by private investment and Canadian government grants, it ran on the web: riders booked there and every driver signed up through a long web form.
Driver supply was the real bottleneck. Interviews and funnel data showed where it broke:
- An 8-step web sign-up with 40+ typed fields and 8 document uploads.
- Licence, work eligibility, driver’s abstract, background check, vehicle ownership, insurance and safety, with rules that differ by province.
- No save and resume, so most drop-offs happened at document upload.
- Drivers work planned shifts, so one no-show breaks a trip between cities.
- Pay questions were the top support topic.
- Joined as a product designer next to Alejandro Delgadillo, who led design. When he moved on I became the only designer, owning design end to end, working directly with the product manager.
- Options first. Early on, Alejandro and I each designed our own version of key screens, and we chose the stronger one. The team, Alejandro included, picked my visual direction as the style for both apps.
- Drove the driver side from the start: onboarding with OCR and automated checks, shifts and earnings. Co-created the Rider app and the shared design system.
- Worked directly with product owners, a project manager who had previously managed product at Uber, and company leadership, turning business goals into flows, prototypes and specs that engineering could ship.
- Cross-functional team: product owners, the project manager, two product designers, iOS, Android and backend engineers, and QA.
Riders book, track and pay. Drivers onboard, plan shifts, take rides and get paid. Both apps share one system and one ride model, so every state a rider sees (searching, driver on the way, arrived, finished) has its pair on the driver side.
Hypothesis. Sign-up data showed most drivers dropped off while typing data and uploading documents. My idea: if drivers photographed their documents and the app filled in the fields from them, far more would finish sign-up. In a daily with the team we checked it was feasible, since OCR already worked in other parts of the app, and it went into the plan. It worked: completed sign-ups went from 31% to 64%.
- One checklistThe web form became a document checklist, with progress saved after every step.
- Scan, don’t typeDrivers photograph their licence and documents.
- Data pre-filledOCR and public registries fill in the data; the driver only checks and confirms.
- Clear review statusA visible status replaced “we’ll email you”.
- A daily goal in hours, with progress for every shift.
- Own availability: preferred, unavailable and flexible time slots.
- “Can’t make it?” with a reason, so dispatch can re-plan supply instead of discovering a no-show.
- Personal goals with a guaranteed bonus for reliability: hours online, shift completion, acceptance rate.
Weekly earnings, instant withdraw, auto payouts and achievements made pay easy to understand and turned retention into a product feature.
- Route and fixed priceThe fare is clear before booking.
- Live matchingRiders see the search and estimated pickup time.
- Driver on the wayCar, plate and driver details with live ETA.
- One-tap tipPreset amounts or a custom tip.
- Scheduled tripsBook longer rides between cities ahead of time.
One visual language for both apps: a Google Sans type scale, a Flame Red core palette with 30 colour styles, 170+ icons and components with every state. 320+ screens across 80+ feature flows, from sign-up to payout.
First 6 months after launch:
- 2× completed driver sign-ups (31% → 64%), with 70% less manual input.
- Time to submit documents: 35 → 9 minutes.
- +38% active drivers, −24% driver churn over 90 days.
- Shift no-shows −32%, with 94% of shifts covered on time.
- Average time to book a ride: 2:40 → 0:55, and +18% more requests turned into completed rides.
- −41% pay-related support tickets.
How we measured: in Amplitude over the six months after the new apps launched. Results reflect the work of the whole product team.
Uride moved from a web-only service to native apps in both stores. The shared design system let the team ship new flows without new UI and supported the expansion to South America. In 2025 Uride ranked #34 in the Deloitte Technology Fast 50 Canada, with 883% growth over four years.
- Rider app
- Driver app
- iOS and Android
- OCR onboarding
- Design system
Make the design system the context for AI from day one. Uride was designed before AI tools were mature, so 320+ screens across 80+ flows were drawn by hand in Figma. Today I would keep the system as tokens and components in code and use Claude Code to build paired rider and driver states as working prototypes, so the team could test flows on a phone sooner and spend less time drawing screens.