Live demo · pre-launch
An AI companion that knows where its authority ends.
Try the live demo
Tracking is solved. Trust isn't.
Pregnancy apps are good at telling you your baby is the size of a peach. They're weaker when you have a real question at 11pm, and the stakes are unusual: most questions are harmless, but a few symptoms (reduced movements, a severe headache with vision changes, heavy bleeding) need a phone call within the hour, not a reassuring paragraph.
Health data is also about as sensitive as data gets. An app that wants people to tell it everything has to earn that by collecting as little as possible.
So the brief: make everyday questions easy, make the dangerous ones impossible to miss, and keep health data on the phone.
Six bets that shaped the product
Safety runs before the AI
Every message goes through a fixed red-flag check first. A match gets a call-now card with the country's numbers, whatever the AI says next. The model can add context but can never downgrade an escalation.
Private by default
Check-ins, moods and AI chats stay on the phone. The server sees a random ID, pregnancy stage, country and active days, and users can switch that off or delete it.
Local from day one
UK, Canada, US and Nigeria each get their own emergency and advice numbers, so "who do I call" is never generic.
Answers with context
Ask Moma knows the week, the stage and recent check-ins, so "is this normal?" gets an answer about week 31, not pregnancy in general.
Humans in the loop
A "talk to the Moma team" chat feeds a support inbox. Red-flag messages jump to the top as urgent.
A short first mile
Onboarding asks two things: due date and country. Every extra question before the first useful moment loses people.
Designing for the 1% case
The red-flag layer is plain pattern rules, deliberately boring and fully testable, written for how people actually type: "baby hasn't kicked since this morning", "soaking a pad an hour", "my face is puffy and I'm seeing flashing lights".
A set of 24 test messages, a mix of real emergencies and everyday questions that only look scary, runs on every change. A missed emergency is never acceptable; an occasional over-cautious card is.
The same set runs in the admin portal's safety lab, so anyone on the team can test a phrase and see exactly which rule caught it.
From first scan to first months
| For parents | For the team |
|---|---|
| Week-by-week journey and a daily check-in | Daily, weekly and monthly active users |
| Ask Moma, the AI companion, with red-flag checks | Activation funnel and weekly retention |
| Baby movement log and contraction timer | Support inbox: assignment, internal notes, urgent first |
| After the birth: feeds, nappies and recovery | Five access levels, from owner to analyst |
| Chat with the Moma team | Safety lab, app settings and a full audit log |
One activation moment, one guardrail
Moma hasn't launched yet, so these are targets, not results.
- Activation: a first check-in within 24 hours of install, the earliest sign the app is part of someone's day. It's built into the admin funnel.
- Retention: weekly cohorts through week 4. Pregnancy is a long journey, so the bar is every week, not every day.
- Guardrail: zero missed red flags on the test set before any release. That number can block a launch on its own.
Built with an AI agent, judged by me
I built Moma with an AI coding agent. My job was the product calls: what the AI may and may not do, which data never leaves the phone, what the safety tests are, and when something was ready to go live. Nothing shipped without passing through a preview I approved.
It runs on Vercel with a Postgres database, and the AI is called from the server, so no keys sit on anyone's phone.
Before real users
- Clinical review of every piece of content and every red-flag rule with a practising midwife.
- A check on whether the symptom checks make Moma a regulated medical device in each launch country, before claiming anything beyond pointing people to care.
- Growing the test set from real, anonymised questions, and a partner mode so both parents can follow along.
- App store releases once the web version proves activation.
"Being right most of the time isn't enough for an AI product. It has to know where its authority ends."
Got a product, a hard problem, or just want to say hi? Send it over. I read every message.
George Odiana
Open to product roles, side-project collaborations and good conversations about hard product problems.
⚡ 1