Property Management, Maintenance Request Triage & Dispatch

Small to mid-size residential property management companies (roughly 50-1,000 units) still triaging maintenance requests manually by phone and text. Best fit: companies with a dedicated maintenance coordinator whose time is being eaten by manual dispatch, or companies getting tenant complaints specifically about slow maintenance response.

A tenant texts or scans a QR code to submit a maintenance issue with a photo and short description, no login required. The system immediately classifies the request by category and urgency (Emergency, Urgent, or Routine), routes it to the correct pre-configured vendor for that category, and texts the tenant a confirmation with an expected response window. The vendor gets the request by text with a simple tap-to-acknowledge flow, and if they don't respond within the SLA window for that urgency tier, the system automatically escalates to a backup vendor and alerts the property manager.

Everything is tracked on a live dashboard showing SLA status per request and vendor performance over time.

The tenant submission form and SMS flows are the straightforward part of this build. The classification logic needs real tuning: expect to spend time testing it against messy, real tenant language ("water thing under sink," "AC not working good," "lock broken again") rather than clean example sentences, and to adjust the default-to-higher-urgency rule so it doesn't over-trigger Emergency tier on genuinely minor issues just because the description was vague. The vendor response flow needs to stay tap-based rather than free-text, since vendors on job sites won't reliably send structured replies, and free-text parsing of vendor responses is a common place this kind of build gets flaky.

Common follow-up requests once the base build works: a way for the property manager to configure SLA windows per property (some portfolios want stricter internal targets than the industry-standard tiers), a tenant satisfaction rating prompt sent automatically after a request is marked completed, and a simple owner-facing monthly report generator built off the same data the dashboard already tracks.