Shawn Mervin

Private mortgage lending software that shows its working.

I built and operate Lend Pilot — pipeline, compliance and borrower portal software for private and alternative mortgage lenders, syndicators and the brokerages that work with them. Canadian math, done properly, and a file a regulator can follow.

Built for how a private desk works

Lead intake to discharge.

A pipeline, not a spreadsheet

Seven stages from lead intake to discharge, with checklists that attach themselves at each stage and projected-funding forecasting.

B-20 qualification in deterministic code

GDS/TDS and the stress test calculated with Canadian semi-annual compounding written into the code, with an underwriting readout that reconciles stated income and shows its working.

A borrower portal with no account to create

Tokenized links for the application, consent capture, document uploads, published offers and live self-serve affordability.

E-signature with a real audit trail

SHA-256 document fingerprints, IP and timestamp capture, and a completion certificate.

Syndication and offer comparison

Compare lender offers side by side and split a deal across investors.

AI on a short leash

Borrower documents — pay stubs, NOAs, statements — are read by AI so nobody retypes them. The model never does the arithmetic.

The platform

Built, shipped, and still running.

Lend Pilot

Multi-tenant SaaS for private and alternative mortgage lenders, syndicators, and the brokerages that work with them.

loans.supersonicrentals.ca

The Lend Pilot landing page, headlined The operating system for private mortgage lending, above a preview of the deal pipeline.
The product front door — pipeline, compliance, syndication, calculators.

A private lending desk is a pipeline problem wearing a compliance costume. Lend Pilot takes a deal from lead intake to discharge across seven stages and qualifies it against B-20 GDS/TDS rules.

  • Kanban pipeline with stage checklists that attach themselves, plus projected-funding forecasting
  • Tokenized borrower portal — application with consent capture, document uploads, published offers, live affordability
  • E-signature with SHA-256 document fingerprints, IP and timestamp trail, and a certificate
  • Lender offer comparison with Canadian semi-annual compounding and per-deal syndication
  • AI extracts data from borrower documents; every dollar figure stays in deterministic code

Result — the desk runs the whole file, lead intake to discharge, and the borrower never makes an account.

Questions people ask first

Straight answers before the first call.

Does AI make lending decisions?

No. AI extracts data from uploaded documents. Qualification, GDS/TDS and the stress test are deterministic code — in a regulated business, a number a model guessed is a liability.

Do borrowers have to create an account?

No. Each borrower gets a secure tokenized link to apply, upload documents and see offers.

Can it carry our brand?

Yes. Every lender runs in its own tenant with white-label theming.

Can you build something custom instead?

Yes — Lend Pilot can be extended, or the same approach applied to a system built around your own process.

Other work I take on

Running your book on spreadsheets and email?

Tell me how deals move through your desk today. If I'm not the right person for it, I'll say so and point you somewhere better.