Thousands of mortgage scenarios are generated, priced and pruned for every file. The results are fully auditable. The method that produced them is protected.
IntelliRate is built so that the reasoning is always reproducible for a client, an auditor or a regulator — while the machinery that produced it stays inside the platform.
A clean comparison of the shortlisted options with the dollar difference stated plainly, and the assumptions behind each figure written out in language a borrower can read.
A timestamped record of every option surfaced, the data the recommendation rested on, the programs excluded by panel, and the specific guideline that disqualified each rejected scenario.
The weighting and interaction logic, the scenario generation and pruning strategy, the approval-likelihood scoring and the lender behaviour models. Outputs are designed so the model cannot be reconstructed from them.
Publishing the taxonomy builds trust with borrowers and regulators. Publishing the model would let any competitor replicate eight years of Canadian lending behaviour in an afternoon.
A recommendation you can't explain to a client is worthless, and a model you can hand to a competitor is worth even less. IntelliRate draws that line deliberately: every figure behind a recommendation is auditable, and the machinery that produced it is not published.
Everything a broker needs to defend the advice, and everything a regulator or client could reasonably ask to see.
Held under NDA, never exported, and not reconstructable from the outputs. This is the part that makes the answer better than a spreadsheet.
In practice: a client sees a clean two-column comparison and a dollar figure; your compliance file holds the complete reasoning chain; and nothing that leaves the platform lets anyone rebuild the engine behind it.
We will go as deep on methodology as a mutual NDA allows — and tell you plainly where we stop.