GLP-1 companion · 0 to 1

Kova

Problem
Most people starting a GLP-1 medication stop within the first year.
Decision
Starting a new business from zero, I framed the product around one outcome: keeping people on treatment in the weeks between doctor visits. That meant solving for the reasons they stop — side effects, adherence, nutrition and muscle loss.
Outcome
Live on the App Store and Google Play just over three months after the idea was first framed. Discovery and delivery ran together, with AI-built prototypes tested with real users and a continuous integration pipeline taking what worked to production.
Titration week view prompting the user to log today's shot
Doses history with estimated medication level over seven days
Symptoms check-in with severity levels
Escalation screen recommending the user contact their clinic
Educational program lessons for the initiation phase

Context

Most recently I led the launch of Kova, a companion product for people using GLP-1 medications.

trykova.comKova on the App Store

What the real problem was

The opportunity is not subtle. Millions of people are starting and taking these medications globally, however a large share stop within the first year. Almost none of the support around medicated weight addresses what actually makes people stop — side effects they were not prepared for, nutrition that has to change as appetite falls and the gap between a prescription and daily life. Customers told us they had two main concerns, starting GLP-1s making the right decision and understanding if they were correctly following their treatment whilst managing any side-effects. 

The decisions worth reading are the exclusions

A first version that deliberately left out several obvious features in order to prove that support between visits really helped whether people stayed on treatment and was of high value to customers. This is a clearer demonstration of product judgement for a zero-1 launch, critical MVP scoping, rapid prototyping and an AI native CI pipeline. 

What happened

Launched MVP within 3 months from concept to production. Using AI first pipeline for design, prototyping, user testing, infrastructure and production code - enabled the business in evaluating and scaling unit economics in this new domain. From first few weeks the product doubled in user engagement week on week.