AI-Assisted Property Operations and Maintenance Triage
A property manager receives maintenance requests through email, messaging, calls, and a service portal.
01Challenge
A property manager receives maintenance requests through email, messaging, calls, and a service portal. Categories are inconsistent, duplicate tickets are common, urgent issues compete with routine requests, and staff spend time searching leases and building manuals.
02Why it was difficult
- Tenant descriptions may be informal, incomplete, multilingual, or ambiguous.
- One incident may be reported through several channels.
- Urgency depends on safety and building procedures, not wording alone.
- Documents must respect building and user permissions.
03Solution architecture
Combine request intake, classification, duplicate detection, retrieval over approved property documents, and maintenance routing. Identify likely issue type, location, urgency cues, and missing information. Escalate potential safety issues to staff.
- Connect approved intake channels and extract building, unit, asset, and issue type.
- Link likely duplicate reports to a shared incident.
- Retrieve procedures from permission-controlled documents.
- Route tickets using urgency rules, trade, location, availability, and service levels.
- Offer draft responses and track reassignment, resolution, and repeat incidents.
04Implementation and validation
Evaluate category precision and recall, duplicate detection, and urgent-case escalation using historical tickets reviewed by staff. Test permission boundaries, outdated documents, conflicting clauses, and incomplete requests. Begin with staff-approved suggestions.
05Outcomes we measure
- Time to correct assignment.
- Ticket reassignment rate.
- Duplicate and repeat-incident detection.
- Document retrieval time and resolution time.
Key lesson
Property automation needs reliable asset records, permission-aware documents, and facilities rules---not just text classification.
Have a problem worth testing?
Tell us about the workflow, decision or research question. We'll tell you honestly whether AI can help and how we'd find out.
Research-led AI, from clinic to company.