Our own rental portfolio: AI-assisted tenant screening and deadline tracking

The challenge

We own a portfolio of houses and condo units across Ontario (the Ottawa area, the GTA, the Hamilton area and elsewhere) plus a furnished condo in Miami. Our property manager handles tenants day to day, and a legal adviser handles tribunal matters. The pressure points:

• Application fraud: edited pay stubs, altered or third-party-generated PDFs, and credit reports whose addresses don't match the application.
• Paper in every direction: property-tax letters, utility bills, insurance renewals, condo notices and vendor invoices across many properties and two countries.
• Deadlines: renewals, tax instalments, insurance expiries and tribunal dates are easy to miss when they live in inboxes and spreadsheets.
• Follow-ups: tenant, vendor and condo-board messages need a quick, consistent and documented reply.

What we built

1. Fraud checks on rental applications. Every applicant package goes through the same protocol:

   • An inventory of every document.
   • PDF metadata forensics on every PDF, covering creation and modification dates, producing software and author fields.
   • Cross-checks of names, employers, income and dates between documents.
   • Every address on the credit report compared with the application.
   • An independent background and credit check through a tenant-screening service.
   • Searches of public tribunal and case-law records.
   • A licence check on any realtor who brings the applicant.

   The output is a risk report, never an approval. The owner decides, and records objective reasons in a decision note with a human-rights-code checklist. The decision note holds no ID, SIN or credit data.

2. Document support. Bills, tax letters, insurance and condo documents are organised by property in shared folders, with amounts and due dates tracked in the portfolio spreadsheets. The AI agent helps review documents and pull out the key details.

3. Deadline register. Renewals, insurance expiries, tax and utility due dates and tribunal dates are kept in one register. Each week it is reviewed for anything due in the next 14 days.

4. Tenant and vendor follow-ups. Replies to tenants, condo boards and vendors are drafted in a consistent voice. Our property manager reviews and sends them, and emergencies go straight to the owner.

Results

• Every rental application now goes through the same forensic protocol, so screening no longer depends on who happens to review the file.
• Each decision rests on a documented risk report, and the owner keeps a written decision note with objective reasons.
• Upcoming deadlines across the portfolio are reviewed together each week instead of being scattered across inboxes.

Tools used

WebKMS Document Automation · Executive AI agent · AI-assisted application screening (now a product, Landlivo) · Google Drive and Sheets · a tenant-screening and credit service · public tribunal and case-law databases · the realtor licensing registry.