— Case Study · Joyride

A 9-month MVP, built in 30 days by one person.

The estimate was three engineers and nine months. I handed the tools to the product manager instead — and shipped the MVP, plus additional features, in 30 days for $100 a month.

Situation

A rideshare where drivers keep 100%.

There was an opportunity to build a new kind of rideshare platform — one where drivers keep 100% of the fares. The level-of-effort estimate for building the MVP was three engineers and nine months.

Task

Compress the cycle time.

Speed up the build cycle using Claude Code, a leading AI coding tool.

Action

One product visionary, plus AI.

I put the tools directly in the hands of the product manager who had developed the vision for the product and the business — and took the traditional build team out of the loop entirely.

What I changed

Empowered the product manager with Claude Code and Claude design.
Subscribed for tokens on Anthropic at $100 per month.
Removed engineers and designers from the development process.

Inside the product

What 30 days actually shipped.

Joyride driver home screen showing online status, a live map, and hot spots with demand multipliers. — The driver goes online — live demand, and where the hot spots are.

Joyride bid screen showing the driver's bid amount, a transparent fare breakdown, and the ride details. — Drivers name their price — a transparent fare breakdown, and a bid the driver controls.

Joyride rider screen comparing available drivers by price and rating, each with a Book button. — The rider compares the bids — sort by price, wait, or rating, and book the one you want.

Every driver has a profile — rides completed, rating, vehicle, and who they are.

Joyride match found screen showing the matched driver, the agreed price, a confirm-ride button, and a link to additional driver options. — Match found — confirm at the price you agreed, or open the other driver options.

Joyride rider tracking screen showing the driver en route, with a warning that the rider is away from the pickup point. — Live tracking — pickup time, ETA, and a warning if the rider drifts from the pickup point.

Joyride rider screen confirming the rider is at the pickup location while the driver is en route. — Geofenced pickup — the app confirms you're in the right spot before the driver arrives.

Joyride earnings report showing total payout, tips, active hours, and a weekly earnings chart. — The driver gets paid — total payout, tips, and per-active-hour stats.

Joyride rider feedback screen showing an overall rating, category ratings, and ride-by-ride reviews. — Rider feedback — anonymous ratings by category, ride by ride.

Result

30 days, 96% less cycle time.

One part-time product visionary completed the MVP build — plus additional features — in 30 days. Cycle time fell 96%, from 2,268 hours to 84. Along the way I developed an entirely new process for AI-native product development.

The outcome

30 days to a complete MVP plus additional features — one part-time product visionary.
96% cycle-time reduction — from 2,268 hours to 84.
$100 per month in tokens, in place of a three-engineer, nine-month plan.
A new process for AI-native product development.

A 9-month MVP, built in 30 days by one person.

What 30 days actually shipped.

Let's talk about driving innovation and growth in your company.