NestBoard
Calendar, chores, meals, and meds on one family screen
My family's coordination ran on a calendar app, a chore chart taped to the fridge, a meal plan in a notes app, and a medication schedule that lived in my wife's head. Every app I tried either stopped at the calendar, ran ads, or charged per family member. So I built NestBoard: one shared screen for a household's calendar, rotating chores with allowance, meals, meds, and routines, plus an AI assistant (Robin) that turns "soccer Tuesdays at 4pm" or a photo of a school flyer into calendar entries.
Where it stands: live on the App Store and Google Play, about 40 real households, first paid conversions just starting as trials mature. One flat price ($4.99/mo per household, unlimited members) because charging a family of eight more than a couple felt wrong. No ads, no data selling, kids join without email addresses.
Stack: React + Capacitor for iOS/Android/web from one codebase, Fastify + Postgres on Railway, OpenAI for Robin. Solo, heavily AI-assisted development.
Three lessons that cost me real users:
My www subdomain was a dead mirror for weeks. The app's security policy only allowed API calls from the apex domain, so anyone landing on www.mynestboard.com from an ad saw a perfect-looking site where signup silently failed. Found it by accident while checking a blog post. If your marketing and app share infrastructure, test every hostname you serve.
Google refused to index my content as "Soft 404" for a reason I'd never read about. My blog posts loaded content by JavaScript, and Google's renderer caches fetches by URL while ignoring headers. The page URL and its data fetch shared a URL, so Google's renderer fed cached HTML to the JSON fetch, the parse failed, and the rendered page showed an error state. Fix: server-render the content and give data fetches their own URL. Same-URL content negotiation is a trap.
A real user's first AI message exposed a hallucination. She asked Robin to "remove burgers from meals this week." Robin had no meal tools at all, and instead of saying so, it confidently replied that no meals were scheduled while burgers sat right there on her plan. Worst possible first impression. We shipped full tool coverage across every feature within days, plus a prompt rule that Robin admits what it can't do. If you're building AI features: the model will fake competence exactly where your tool coverage ends.
Next up: per-household email addresses (forward the dentist confirmation, it lands on the calendar) and watching whether honest SEO content (we publish comparisons that recommend competitors where they're genuinely better) can out-rank content farms.
It's at mynestboard.com. Happy to go deep on any of it: Capacitor monorepo pain, COPPA for kid accounts, AI tool design, or bootstrapped family-app economics. What would you want to know?
1 Comment
I like that you started with the reality that families don't use one app—they use a collection of workarounds held together by memory and group chats.
The interesting part isn't combining calendars, chores, and meals into one place. It's reducing the amount of coordination families have to do just to stay on the same page. That feels like a much more meaningful problem to solve than simply adding more household features.