SPEC.CASES / 01
The work, in depth.
Long-form breakdowns of the product and growth work I'm proudest of: what the job was, how I ran it, and what actually happened. Each one is a real project, with the numbers and the lessons attached.
Every case study.
Key impact
7 sources unified · used daily
I built Drops an internal AI agent that turns scattered product context into answers the team can trust.
How I built Drops one AI agent wired to every source our product context lives in: four modes, 17 on-demand skills, and a citation for every answer.
READ →Key impact
+15.6% engagement
A spaced-repetition engine that surfaces each word just as you start to forget it.
I rebuilt Drops’ spaced-repetition review around a memory-fading model that surfaces each word as it starts to slip: +15.6% engagement, +37% for new users.
READ →Key impact
+20% ARPU
A push engine that stays relevant as users change, and fresh as the months pass.
Rebuilt Drops push as a lifecycle engine: 18 scenarios × 5 user stages, with copy that rotates monthly. +24% reactivation, +22% conversion, +20% ARPU.
READ →Key impact
Validated growth upside
Replacing the 5-minute limit that told most of our users to leave.
A high-risk bet to replace Drops’ 5-minute paywall with a freemium model that grows engagement, rankings and revenue. De-risked in stages, then parked.
READ →