Build in public

Every landlord has a version of this story. Mine involves a boiler, a February, and a phone that would not stop buzzing.

Property management, done right, is mostly a job of showing up on time for small things: the rent reminder on the 3rd, the plumber follow-up, the lease renewal you should have started sixty days out, the tenant text at 11pm that can't tell you itself whether it's an emergency. None of it is hard. All of it is relentless. After nine years I could do the job in my sleep — which was convenient, because the job kept happening while I was trying to.

The industry's answer to this is software that organizes your to-do list beautifully and then hands it back to you. I had the to-do list. I didn't want a prettier one. I wanted an employee.

For a small building, full-service management can be difficult to justify, while a collection of point tools still leaves the owner coordinating the work. I had been coding since childhood, and I have managed residential rental property since age 13 and have owned since 21. In 2026 I started building software to make routine coordination more structured without assigning consequential judgment to a model.

The product I wanted, which Aptoria has planned, would take on routine work and ask for judgment at defined stop points, such as scheduled rent reminders, payment attempts, maintenance triage into dispatch proposals, and notice drafts held for review. Today Aptoria covers the records underneath that work: properties and units, the rent roll and balances, a monthly reconciliation, and notice templates. Its action history is an operational review aid; it is not a complete legal record or proof that an external provider completed an action.

Then I did the only honest next thing: I put my own building on it.

The planned first pilot is a turnover in my Brooklyn building. If the unit, tenant timeline, provider setup, and required approvals are ready, Aptoria will help coordinate supported listing, inquiry, showing, application-assembly, lease-draft, and move-in steps. I will retain every applicant decision and any external action that requires owner or professional review.

I intend to publish measured results after each completed reporting period, with the denominator and method beside them. Until a period closes and the records are reviewed, the scoreboard remains blank; this page does not present planned work as a completed outcome.

Planned one-unit pilot, n=1. A cell fills only after its measurement period closes and the source records are reviewed.

The same disclosure will cover observed errors, interventions, exclusions, and material corrections. Where an eligible action used the soft-commit workflow, I will identify whether it was cancelled during the available window; provider-managed or already-completed actions can follow different rules.

My rental experience shaped the current human-required floor. Supported Aptoria workflows do not autonomously execute defined consequential categories such as eviction filings, applicant denials, or security-deposit deductions. The product may organize records or prepare a draft, while the authorized person retains the decision and any required legal or provider process. That boundary reduces risk; it is not a legal-compliance guarantee.

A few design decisions that came straight from operating a real building rather than imagining one.

I'm building this in public and plan to publish reviewed pilot measurements here. If you own a few units and want to evaluate Aptoria's supported workflows and review controls, you can join the product without treating this planned pilot as proof of your likely result.

If you want to tell me why this will never work, my inbox is open — skepticism from working landlords has already shaped half this product.

This is n=1. One unit, one building, a small sample — not a portfolio-wide average, and I say so on the scoreboard itself. Every number I publish carries its denominator.

There are no pilot results to show yet. A number publishes only after its measurement period closes and the supporting records are reviewed; until then the cells remain blank rather than modeled.

A single building cannot predict another property's outcome. New York City has its own local operating and legal complexity, and each customer remains responsible for verifying the rules that apply to their property.

Aptoria's founder. Has managed residential rental property since age 13 and has owned since 21; coder since age seven; finally both at once.

Is this really your building?

n=1 isn't proof. Why publish it?

What happens when the AI screws up?

Why should a landlord in Texas care about a Brooklyn building?