Bonus
I even got to compose the music for the announcement video
Case study / Revel
Designing the passenger experience inside an all-electric rideshare fleet
At Revel I designed the Pax App, a tablet interface mounted in the back seat of every Tesla in the fleet, turning an untouched surface into a differentiated passenger experience.
Background
A few things to know before diving in.
1
Revel
Revel launched in 2018 as a Brooklyn-based moped sharing platform and expanded into an all-electric Tesla rideshare service operating in New York City.
2
Full Fleet Ownership
Revel owned the entire fleet outright. That gave the company end-to-end control over the rider experience, from the consumer app to the moment a passenger stepped out of the car, in a way Uber or Lyft never could.
3
The Back Seat
Revel installed a custom tablet in the back seat of every Tesla in the fleet. Hardware sourced and mounted by Revel, it turned an otherwise ignored surface into a dedicated passenger screen.
4
First Designer
I joined Revel as the first designer. The Pax App was one of many surfaces I built from scratch, alongside ops tools, the consumer moped app, and the company's design system.
The situation
A blank screen in the back of every car, and no playbook for what to put on it
Revel was betting on brand, on being a fundamentally different kind of rideshare company. The back seat tablet was a chance to make that real. But no one had designed for this context before. Custom hardware installed by Revel, the tablet occupied a surface no other rideshare company had taken seriously: a shared, semi-public screen inside a moving vehicle, used by strangers for 10 to 30 minutes at a time, usually without any clear intent going in. Mirroring the mobile app was the obvious first instinct. That framing didn't hold up.
The problem I set out to solve
How might we use the back seat tablet to create a passenger experience that actually reflects what makes Revel different?
The task
Design a new kind of in-ride experience from scratch, for a surface with no precedent
The brief had three threads: distinguish the product in the market, give passengers something to actually engage with during the ride, and use the surface to tell the Revel story. Achieving all three meant designing something that felt ambient enough to ignore and interesting enough to explore.
The constraints
1
The surface is shared and semi-public. Content needed to work for any passenger, regardless of language, age, or familiarity with the product.
2
Tesla's API access was limited. Many features that seemed obvious in early exploration ran up against hard technical walls.
3
The session is inherently short and unintentional. Passengers are in a car, not actively choosing to engage. Anything too complex or demanding would go untouched.
4
Pairing had to be frictionless. A passenger shouldn't need instructions to connect to their own ride's screen.
5
The screen needed a safe default state: something that looked good and communicated brand even when no one was interacting with it.
The action
A wide net cast, then narrowed through testing and constraint
The research phase started with a clarity/risk matrix: mapping what we knew well and what needed investigation before committing to design direction. Early steps included a competitive landscape sweep, hands-on sessions with clickable prototypes in actual vehicles, and a close look at what Tesla's API would and wouldn't allow.
Early explorations went broad: music integration, weather displays, climate controls, seat warmers, company information, driver profiles, and video content all made it into the first round. Each feature was stress-tested against the constraints: does it work for a stranger with 12 minutes to kill? Does it require setup? Does it break if someone ignores it entirely? That filter eliminated a lot.
Design explorations
The result
The MVP: a coherent in-ride experience built around ambient presence and passenger choice
The shipped product opened with a pairing flow. Passengers entered a code to link their session to the vehicle, then settled into a sleep screen showing the time, local weather, and a subtle prompt to explore. From there, passengers could browse radio stations, check the weather, read about Revel's drivers and mission, control climate settings and seat warmers, or watch company videos.
Every feature was scoped to the context: no accounts, no friction, no feature that required more than a tap or two to reach. The design also included a small easter egg: a hidden screen that surfaced app version info for drivers and a 🍄 unlock moment. A little personality in an unusual context goes a long way.
Outcome
The MVP launched, engaged passengers, and ran into a wall it couldn't design its way out of.
Early engagement data and driver observations were encouraging. Passengers were interacting with the screen, and the anecdotal feedback was positive. The design had real legs. Then a DOT ruling changed everything: removing the front passenger seat (a requirement for the tablet mount configuration) was declared illegal, making the hardware setup untenable. The project ended not because the product failed, but because the physical constraint it depended on was regulated away. I still believe the core idea was sound. The work stands as a template for what a differentiated in-ride experience could look like when the hardware situation allows.