What AI is, and what it is not: a plain-language guide for business leaders
A short guide to what today's AI tools can do well, where they fail, and how to talk about them without hype.
AI for property management: five use cases that can pay back
Five practical places where AI helps property teams today, and what to check before you try each one.
There is a lot of noise about AI. In property management the useful applications tend to be unglamorous: repetitive, text-heavy work that someone has to check anyway. Here are five places where AI can help, with the main thing to watch for in each.
1. Sorting and routing incoming requests
Tenants write in many ways: "AC not cooling", "water on the floor in the kitchen", "my unit is too hot". AI can read these, classify them (plumbing, cooling, electrical), judge urgency and send them to the right team. Watch for: emergencies. Keep a simple rule that anything involving safety is escalated to a person immediately.
2. Drafting replies and notices
AI can prepare a first reply to a tenant, an owner update or a notice, in your tone and in the right language. A team member reviews and sends. Watch for: promises. Make sure drafts never commit to dates, costs or legal positions without a person approving them.
3. Answering questions from your own documents
An assistant that reads your procedures, lease templates and building rules can answer staff and tenant questions and show where the answer came from. This is the RAG approach described in another article here. Watch for: old documents. If the source is out of date, the answers will be too.
4. Summarising long threads, calls and reports
Managers spend hours reading email chains and meeting notes. AI can produce a short summary with open actions and who owns them. Watch for: missed detail. Use summaries to decide where to look, not as a replacement for the record.
5. Spotting patterns in operational data
Repeated failures on the same equipment, units with unusually frequent complaints, or invoices that look different from the usual. AI-assisted analysis can surface these for a person to investigate. Watch for: false confidence. A flagged pattern is a question to ask, not a conclusion.
How to choose
For each idea, ask three questions:
- Is the work repetitive and frequent enough to matter?
- Can a person check the result quickly?
- What is the cost if it is wrong, and who catches it?
The best first projects score well on all three. They save time you can measure, they are easy to review, and mistakes are cheap and visible.
What to avoid at the start
Anything that makes financial or legal decisions on its own, anything that uses sensitive personal data in tools you do not control, and anything you cannot measure. Prove value on small, safe tasks first. The confidence you build will carry the bigger projects.