Ozyn is an AI assistant that does real work. It researches, drives a browser, fills spreadsheets, sends email, and pauses to ask permission before doing anything consequential. It runs today on web, Slack and Telegram.
It does not run properly on a phone. That is the single biggest gap in the product, and you would be the person who closes it.
THIS IS OUR FIRST MOBILE HIRE
You own iOS end to end: the architecture, the code, the App Store Connect account, releases, crash triage, and the decision about what stays web and what goes native. There is no mobile team above you or around you yet — you are the start of one.
WHY THE PHONE MATTERS MORE HERE THAN FOR MOST PRODUCTS
Ozyn's core interaction is an approval gate: the assistant works for several minutes, then stops and asks before it sends an email or spends money. Today that question arrives on a screen the user may not be looking at. A phone with a working, actionable notification isn't a nice-to-have for this product — it's the missing half of the core loop.
THE FIRST THIRTY DAYS
We want the app live on the App Store within your first month. That is a deliberately hard target and we would rather say so plainly than dress it up: it means shipping the shell we already have, well, rather than building the app you would design from scratch. The native rebuild comes after, informed by real usage.
• Week one — Revive the shell, decide which of our two web skins it wraps, and put a TestFlight build on every phone at Tipstat, rough edges included.
• Week two — Sign-in working on every path: phone OTP, Google and Microsoft native bridges, and Sign in with Apple, which App Store review requires of us because we offer third-party login. Protected areas and the Energetic Island handled properly.
• Week three — APNs live end to end: approve or reject a pending agent action from the lock screen, exactly once. Live progress for twenty-minute runs that survives backgrounding.
• Week four — Through App Store review and publicly l