Mortgage lending should be faster, smarter, and less painful
Mortgage lending is one of the largest credit categories and one of the most frustrating borrower experiences, and it is worse for the self-employed. For traditional employees, the process is relatively simple: pay stubs and an employment letter. Self-employed individuals must show tax returns, bank statements, and client letters to prove they make enough money.
Existing automated underwriting systems are heavily biased toward predictable, salaried paycheques. Self-employed borrowers are frequently routed to alternative lending programs carrying larger down-payment requirements and rate premiums. Being denied a home loan because one is self-employed carries a heavy emotional and mental burden. When you build a successful business, you expect society to view it as an achievement, but the housing market often treats it as a liability.
I experienced this problem first-hand five years ago, when I was self-employed and struggled to convince lenders that I was a credible borrower.
Why now?
Converging technological and market forces make this the moment to overhaul an archaic system.
Roughly 2.7 million Canadians (about 13% of the workforce) are self-employed (StatCan), alongside some 16.6 million Americans (BLS). Yet their share of mortgage originations runs well below their share of the workforce: the market systematically under-lends to the people most likely to own businesses. Research in comparable markets shows self-employed applicants are over 40% more likely to be rejected by every available lender, and face a third fewer lender options than employed borrowers (Mortgage Broker Tools, 2026). BDC estimates that unlocking the growth of Canada's self-employed could add $24 billion a year to GDP, but that growth runs through their ability to access credit as easily as they create value.
AI collapses the mechanical work in a complex file from days to minutes. The opportunity isn't AI replacing underwriting judgment; it's AI-era infrastructure built to the standard that judgment requires: explainable, auditable, repeatable. Costs stop being passed to borrowers as premiums, approvals accelerate, and fewer exhausted applicants abandon mid-process or close with a competitor.
Introducing Vouched
Introducing Vouched: a tool for brokers and lending teams that handles the interpretive work self-employed files demand, including add-back calculation, affordability stress testing, lender narrative generation, and document verification. It works alongside the submission platforms your firm already uses. Vouched is the preparation layer for complex files: collect, verify, and package borrower documentation into complete, underwrite-ready files.
Borrowers are saddled with complex paperwork, hunting for missing financial records that cause delays and fatigue. Brokers and lending teams, for their part, spend hours on emails and calls chasing missing documents and verifying that what arrives is genuine and unaltered. Vouched removes the redundancy in that workflow. It doesn't replace the judgment of a broker or an underwriter; it clears everything standing between them and that judgment. The document chasing, the add-back math, the verification legwork: all handled. What's left is the decision, made faster and with a more complete picture.
Built for the files everyone else avoids
For borrowers, Vouched replaces the document scramble with a single flow: upload what you have, connect what you don't (bank accounts, tax records), and the file assembles itself, complete and consistent, in hours instead of weeks.
For brokers and lenders, Vouched does the interpretive work self-employed files demand. Income history reconstructed from T1s, NOAs, T2125s, and corporate T2s. Add-backs calculated deterministically, every figure traceable to the rule and the document that produced it. Business stability and consistency checks run across every source before the file moves. Regulators expect underwriting decisions on complex files to be explainable, and OSFI's expectations for model risk and income verification are only tightening. A deterministic engine with a complete audit trail isn't just faster than manual review; it's the version of this work that survives scrutiny.
Every lender underwrites self-employed income differently: which add-backs count, how many years to average, what stability looks like. Vouched's qualification engine encodes these rules deterministically, configured to your guidelines, applied identically to every file, with every calculation traceable to the rule that produced it. No black boxes, no model drift. The same logic your underwriters apply, executed in seconds instead of days.
When documents are missing, Vouched chases them with automated follow-ups sent under your name, so files never stall in an inbox. Every document is checked for consistency: income figures reconciled across tax filings, bank statements, and stated amounts, with discrepancies surfaced early. Brokers place stronger deals. Lenders decision cleaner files. Everyone stops chasing paper.
Where we're headed
Vouched is becoming the orchestration layer for complex mortgages: CRA-authorized income verification, IRS tax transcripts, payroll verification, bank-sourced statements, and an assistant that answers borrowers instantly, so document fatigue stops killing applications.
Where we begin
Vouched works with the most ambitious lenders and brokers to simplify their workflows and eliminate redundancy. It removes hours of manual labour across document processing and verification, and supports underwriting on the most complex self-employed files.
Our goal is simple: ensure that mortgage brokers and loan officers have every detail they need to make an informed decision on a self-employed file. We've built infrastructure that eliminates the inherent redundancy in the complex-file origination workflow, giving teams their time back for the work that matters most each day.
Today, that means the best-prepared self-employed files in the market. Tomorrow, it means the rails they run on: from first document to final decision.