Solo-built a life-admin app entirely from my phone, launching on Product Hunt in 3 days
Hey IH — I'm Chris, a veteran and solo founder. Wanted to share what I've been building: Actually Done, a life-admin app that keeps bills, subscriptions, documents, passwords, and family records in one place instead of scattered across a dozen apps.
The interesting part for this community: I built the entire thing using Replit's AI agent, working entirely from my phone, with no formal coding background. Went from idea to a live 36-section app with Stripe billing, encrypted password vault, and OAuth in a few weeks.
Launching on Product Hunt July 21. Happy to answer questions about the build process, working with an AI agent as your primary dev tool, or anything about going from zero to shipped as a non-technical founder.
Building from a phone is impressive, but the more interesting lesson may be how you decided when to trust the agent and when to stop.
Passwords, billing, family records, and OAuth all have very different failure costs from ordinary UI work. For a non-technical founder, a simple escalation rule could be: let the agent handle reversible implementation details, but require independent evidence or expert review before accepting decisions that expose private data, money, or irreversible changes.
Did you develop any explicit rule like that during the build, or was it mostly decided case by case?
With 36 sections, what happens when someone wants to leave? A clean full export and recovery flow might matter almost as much as onboarding for a product holding that much personal data.
The Product Hunt page structure matters more than the phone-build story here. Personal life-admin tools live or die on whether the first five minutes prove it's not another app you now have to maintain. I'd lead with the messiest real example, a bill plus a password plus one document captured and found again in under a minute, that's the moment that converts skeptics, more than how it was built.
The 36 sections make breadth the headline, but passwords, documents, and billing in one product create one large trust boundary. I’d launch with one compartment, such as expiring documents and reminders, plus a visible export and recovery drill. If users won’t entrust that narrow slice, more categories won’t solve adoption.
Building something that holds passwords and family records entirely from a phone using an AI agent is impressive, but that's also exactly the kind of app where trust has to be earned before anyone hands over real data. The security questions in the comments aren't nitpicking, that's the actual first hurdle for this category of product. What's the very first thing a new user does that makes them trust it enough to add something sensitive?
Building and shipping a product entirely from your phone is impressive, but the part I'd be most interested in isn't the AI workflow. It's how you're validating something as personal as life admin.
When one app stores passwords, financial details, and important documents, trust becomes part of the product. I'd spend as much time communicating backup, encryption, and data recovery as the feature set, because those are the questions that determine whether people are willing to move their lives into a new app.
The interesting opportunity isn't proving that an AI agent can build a production app—it's becoming the one place people trust to manage life's administrative complexity. I'd keep validating whether users stay because everything is centralized or because they finally feel in control of responsibilities that used to be scattered across multiple tools. That's a much more durable position.