Building Sunny Seed we had to decide this feature by feature, and the answers turned out not to be consistent.
On the phone: WHO growth percentiles - the LMS tables ship inside the app, it's arithmetic, and a server round-trip to divide two numbers would be absurd. Photo tagging for smart albums, via Apple Vision on iOS and on-device ML Kit on Android. And dictation on iPhone, which is transcribed locally and never becomes a file.
In the cloud: the AI assistant. Not because of model quality - because it needs months of a child's history in the context window to say anything useful, and that's the part a phone can't hold at a speed anyone would tolerate. Decode speed gets all the attention in these discussions, but prefill on tens of thousands of tokens is where local inference quietly stops being practical.
The pattern that emerged for us: narrow, purpose-built models go local. Anything that has to read your whole history stays on a server.
Hi Product Hunt 👋
I'm Roman. I've been building mobile apps for 15 years, and I'm a father of a little girl.
Sunny Seed started as our own problem: my wife and I had notes about our daughter scattered
across three apps, a paper notebook, and a chat with grandma. Nothing was in one place, and
by the time you notice, half of it is gone.
So I built the thing we needed. My wife is a children's book illustrator - every illustration
in the app is hers, none of it is stock.
Two decisions I'd genuinely like your take on:
On-device where it counts. WHO growth percentiles are computed locally from LMS tables
that ship inside the app - no server round-trip to tell you where your child sits on the
curve. Photo tagging for smart albums also runs on the phone (Apple Vision on iOS, on-device
ML Kit on Android), and dictation is transcribed on-device without the audio ever becoming
a file. Our backend is in the EU. For a product holding data about a baby, that felt
non-negotiable - even though it made some features harder to build.
The AI doesn't play doctor. The assistant answers everyday questions about sleep, feeding
or solids, but for anything medical it consistently points to a pediatrician. That's wired
into the product, not a line in the small print.
The part we care about most isn't a feature list: write a few entries during the month and
they turn into a letter about that month. Not a chart - text you'll actually want to reread
in five years.
What it isn't: we're early. Both iOS and Android are live, but young. If you try it and
something feels off, I'd much rather hear it here than read it in a review.
A question for the parents here - what did you actually keep from the first year, and what
did you stop tracking after week two? Genuinely curious, because our own answer surprised us.