Property Management Case Study: AI Maintenance Intake
How a property manager with 640 units stopped fielding maintenance calls at the front desk, with an AI that logs every request, sorts emergencies from routine work and dispatches the right vendor.
Overview
The client manages about 640 residential units across Los Angeles County with a two-person front desk. Maintenance requests came by phone, email and text, and after 5 PM they went to a voicemail that the on-call manager checked when she could. A leaking water heater at 9 PM and a squeaky door got the same treatment: a note, and a call back tomorrow.
Gain Rhino replaced the intake with an AI that answers every maintenance call, text and email, day and night. It identifies the tenant and unit, asks what is wrong, takes a photo if the tenant has one, and creates a ticket in the CRM. Emergencies, defined by the property manager, page the on-call manager and dispatch the right vendor from the approved list immediately. Routine requests are scheduled with the vendor and confirmed to the tenant by text.
Research
We logged 30 days of maintenance contacts: about 520, with 38 percent outside office hours. The front desk spent about half its day on maintenance calls, and after-hours emergencies waited an average of three hours for a callback. Tenants called back repeatedly for status, which added to the load.
The property manager wrote the emergency rules: active water leaks, no heat or hot water, electrical hazards, lockouts, anything involving safety. Everything else is routine. We connected the approved-vendor list by trade and area so dispatch picks a vendor that actually covers the building. Tenants get status texts at ticket creation, vendor confirmation and completion, which is where most of the repeat calls came from.
Result
Every maintenance request is now logged the moment it arrives, sorted correctly, and either dispatched immediately or scheduled with a confirmation to the tenant. The front desk handles exceptions instead of every call. The on-call manager is paged for real emergencies only.
By the numbers (first 90 days):
- Maintenance requests logged as tickets100%, up from about 70%
- Emergency response time after hoursabout 20 minutes, down from 3 hours
- Front-desk time on maintenance callsdown roughly 60%
- Tenant status callsdown about 70%
- Tickets dispatched to a vendor without staff involvement3 in 4
Client identity withheld at their request. Figures reflect the first 90 days after launch.
