Field guide 03 · Property operations
Triage maintenance requests while keeping urgent decisions with people
Classify incoming repair requests, identify missing information, and prepare a response draft while a person confirms urgency, responsibility, access, and dispatch.
The direct answer
Can AI prioritize tenant maintenance requests in BC?
AI can organize requests and flag language associated with urgent conditions, but a person should confirm the classification and act. Emergency repair, health, safety, access, privacy, and tenancy decisions should not be automated.
Good fit when
- Requests arrive through a defined channel
- The team has written emergency and routine triage rules
- A person monitors the queue during service hours
- There is a direct emergency route outside the AI workflow
Keep outside the boundary
- Autonomous emergency classification or dispatch
- Determining legal responsibility or denying a repair
- Generating notices or entering a unit without required review
The workflow
Assist the work. Keep accountability visible.
A coordinator reads emails, texts, forms, and voicemails; asks follow-up questions; decides what is urgent; opens work orders; and updates tenants and vendors.
AI may assist
- 01
Place requests into a review queue by property and issue type
- 02
Flag possible emergency language for immediate human attention
- 03
Identify missing details such as location, access, photos, or active damage
- 04
Draft an acknowledgement and work-order summary
People must retain
- 01
Confirm whether an issue is an emergency and take immediate action
- 02
Apply tenancy rules, lease terms, strata rules, and property procedures
- 03
Decide responsibility, access, vendor assignment, and timing
- 04
Approve communications and maintain the official record
- Written maintenance request
- Property and unit identifier
- Approved emergency criteria
- Property-specific contacts and procedures
- Optional photos supplied by the requester
A review queue showing the original request, possible urgency, missing details, proposed category, draft acknowledgement, and required human action.
30-day pilot
Small enough to inspect. Real enough to learn.
Use historical or low-risk work first. Preserve the original inputs and current-process result so the comparison remains honest.
Set the boundary
Choose one owner, one maintenance-request queue, approved inputs, and a human approval point.
Test side by side
Run a small historical sample through the current process and the assisted process. Keep the original records.
Measure the exceptions
Track correction time, missed facts, escalations, and whether the proposed output was actually useful.
Decide with evidence
Document the result, update the rules, and continue only if the workflow is safer or meaningfully better.
Measure
Count the corrections, not just the speed.
- Time to first human review
- Confirmed urgent requests surfaced correctly
- False urgent flags and missed urgent requests
- Follow-up messages needed to open a work order
- Tenant information exposed beyond its necessary purpose
Stop the pilot if
- The tool misses or down-ranks a confirmed urgent request
- Staff rely on the label without reading the original message
- The workflow slows the emergency path
- Personal information is collected or retained without a clear need
A failed pilot is useful evidence. It may point to a narrower workflow, better data, stronger review, or no AI at all.
Common questions
Clear answers before the pilot.
Can the system decide that a repair is an emergency?
It should only flag a request for urgent human review. The responsible operator must confirm and act under the applicable rules and procedures.
What is a safe first scope?
Draft acknowledgements and identify missing information for routine requests, while routing any possible emergency directly to a person.
Should tenant messages be used to train a public AI model?
Not by default. Review the tool’s data-use terms, minimize personal information, and establish an approved business environment first.
Source trail
What informed this field guide
This is a practical workflow synthesis, not reported ROI or a claim that the use case has been validated for your organization.
- Province of British Columbia — Residential Tenancy BranchRepairs and maintenance
Explains landlord and tenant responsibilities, written repair requests, and the conditions and process for emergency repairs.
- Office of the Information and Privacy Commissioner for BCPrivacy resources for private organizations
Provides BC PIPA resources for organizations that collect, use, and protect personal information.
- Office of the Privacy Commissioner of CanadaPrinciples for responsible, trustworthy and privacy-protective generative AI
Supports necessity, proportionality, transparency, accountability, and privacy review before personal information enters an AI workflow.
- National Institute of Standards and TechnologyAI Risk Management Framework
Provides a voluntary structure to govern, map, measure, and manage AI risk throughout a pilot and deployment.
General operational education only. Confirm applicable legal, privacy, cybersecurity, professional, safety, contractual, and workplace requirements for your organization.
Make it specific
Turn this pattern into your Project Brief.
Define the owner, real inputs, approved output, success measure, and constraints before evaluating a provider.