Weekly property-level signals across the entire Canadian residential market. Match BrightCat's pipeline against your mortgage book and you'll see every listing, every price change, every sale on a property you're financing — 4 to 12 weeks before discharge.
By the time the discharge request arrives, the property has been on the market for two months. BrightCat puts you on it in week one.
A mortgage discharge is a lagging signal. By the time the discharge request lands at the bank, the property has been listed, shown, negotiated, and is days from closing. The retention conversation has already been lost.
For Canadian federally regulated lenders, mortgage retention is a structural problem. Renewals get shopped, refinances go to brokers, and discharges follow property sales the lender didn't know about until it was too late. The signal that the borrower was leaving was visible weeks earlier — on the property listing — but didn't make it into the lender's data environment.
The bank already holds the address, the borrower record, and the loan terms. What it doesn't hold — and historically couldn't easily acquire — is a weekly, property-level view of what's happening on the asset side of the book. BrightCat is the missing layer.
The match key is the property — not the borrower, not the loan number, not the credit file. Properties have persistent identifiers; borrowers move, refinance, and change names.
Six recurring signal types BrightCat fires on properties securing active mortgages.
Documented methodology, versioned builds, per-record lineage. The procurement and model-risk paperwork was anticipated.
BrightCat's pipeline has produced a versioned weekly build every week since 2014 without interruption. Each build carries a build identifier, capture-window timestamps, and per-record lineage tying every signal back to its source observation. For institutions running OSFI Guideline E-23 model risk reviews, Treasury Board Directive on Automated Decision-Making assessments, or internal AI governance audits, this matters: the data layer is documented, reproducible, and auditable.
Production deployments operate under an annual Master Data License Agreement (MDLA) with defined audit rights, AI/ML usage clauses, and lifecycle terms. Sample data evaluation is free and runs under a standard NDA. Full methodology documentation, including the persistent property identifier construction, is available under NDA for procurement and model-risk teams. Methodology proof page →
No portal logins for analysts. No CSV pulls from a vendor extranet. The signals land where the rest of the bank's data lives.
Mortgage portfolio monitoring is the practice of continuously tracking the status of properties securing an active loan book. For Canadian federally regulated lenders, this typically includes listing detection, price change tracking, sold-event confirmation, and discharge prediction.
BrightCat captures property-level listing events weekly across all 10 Canadian provinces. By joining the lender's mortgage book on standardised address and persistent property identifier, BrightCat flags every listing in the book typically 4–12 weeks before discharge.
Credit bureau monitoring captures the borrower side. Property listing monitoring captures the asset side — the property is on the market before the borrower applies anywhere. The two are complementary, but property listing typically leads credit inquiry by weeks.
BrightCat's pipeline is documented end-to-end with versioned weekly builds, per-record lineage, and methodology documentation available under NDA for model risk reviews.
Snowflake Marketplace with Secure Data Share, MCP connector, flat file (CSV / Parquet) to the bank's cloud storage, or hashed match-and-return for institutions with outbound data constraints.
Send a sample slice of your mortgage book under NDA. BrightCat joins, returns the historical signal trail, and you measure the lift on a real portfolio before discussing terms.