Work Sample build
Five AI agents running maintenance, start to finish
A team of AI agents for a property manager with about 140 rental units. Intake, Dispatch, Tenant Comms, Owner Updates and Billing hand each repair request from one to the next, from the tenant’s first text to the owner’s invoice, and bring anything over the limits to a person.
- Built for
- A 140-unit property manager
- Build type
- AI agent workflow
- Typical build
- 6 to 10 weeks
Sample Maplecrest Property Management is a fictional business and the data is made up. The AI text in this demo is pre-written so it runs without an account; in a real build it’s written live, and you choose what goes out on its own and what waits for a person.
Best on a laptop; on a phone it shows one panel at a time. Open it full screen (opens in a new tab)
01 What it does
- Intake reads each tenant request by text, email or portal, sorts it by trade and urgency, and asks for a photo when it helps.
- Dispatch compares vendors on availability, rating and rate, books the best fit and explains why it chose them.
- Tenant Comms and Owner Updates keep everyone informed; Billing matches the vendor invoice to the quote before the owner is billed.
- Guardrails you set: spending over $500, invoices over quote, after-hours calls and unit entry all wait for your OK.
02 Try this
- Watch the activity feed as the agents work.
- Click Simulate new request and follow the handoffs.
- Open Waiting on you and approve the $1,240 water heater.
- Open Agents to see each agent’s rules and limits.
Want one built around your business?
You get one fixed price in writing before we start, you approve every release, and the first call is free.