I lead product for mortgage lending: borrower experience, loan operations, and sales technology across purchase, refinance, and HELOC. More and more of that work is applied AI, like an underwriting agent that verifies income and employment and decides which files need a person.
Every decision has to balance borrower experience, operating cost, credit risk, compliance, investor requirements, and engineering capacity. I bring order to the chaos.
My Superpowers
- Applied AI in lending: I set the strategy and requirements for AI systems that make lending decisions, and I build early prototypes myself to test ideas before engineering commits.
- Product org leadership: I own strategy and roadmap across several pods, and I hire, level, and coach the PMs on my team.
- Growth and experimentation: I run A/B and multivariate tests. That work raised lead conversion 9.7%, worth about $400K a month.
Notable Highlights
- AI underwriting agent: Automated income, employment, and residency verification. Review time fell 50%, conversion rose about 40%, and the work produced $6M in incremental profit.
- Appraisal workflow: Cut appraisal turnaround from 10 days to 3, which let the company grow from 800 to 2,000 rate locks a month.
- Risk qualification: Built an FCRA-compliant risk qualification flow that supported $436M in securitizations and raised retention 7%.
- 0-to-1 lending: Launched a mortgage point-of-sale system and an eClosing platform projected to process $1.5B.
- Compliance beyond lending: Delivered trade compliance reporting for ESMA requirements, and a patient self-service portal that handled 1,200 inquiries a day during the pandemic.
What I’m Building Now
An underwriting agent that reads documents, calculates income, and acts on the result. It advances files and requests missing documents without a person stepping in. A multi-AUS routing strategy that runs each loan through more than one automated underwriting system to get the best outcome. A home equity product line that’s funded and adding partners.
I also built a tool that pulls tax-delinquent properties from county records, ranks them as tax lien purchases, and runs basic title searches.
What Matters Most
I’ve spent my career in regulated industries: mortgage, fintech, and proptech. Every feature touches compliance, capital markets, or investor requirements, and mistakes are expensive. I like the focus those constraints force.
I hire people who are better than me at their jobs and give them room to work. I coach more than I direct, and I push back on engineering or the business when either one is wrong.
The decisions I care most about are the ones where every option costs something, like borrower experience against operating cost, or speed against regulatory risk.
Let’s Connect
If you’re working on hard problems in lending, fintech, or applied AI, I’d like to hear about them.
| 📧 jonathan@jvasquez.net |
Case Studies
-
AI Agent
Automating Loan Underwriting with an AI Agent at MegaLoans
An AI underwriting agent that verifies income, employment, and residency cut review time 50%, scaled capacity to 1,000 loans a month, and lifted conversion about 40%.
-
Automation
Replacing Manual Appraisal Tasks with Automated Data Collection
Cutting appraisal turnaround from 10 days to 3 let MegaLoans grow from 800 to 2,000 rate locks a month.
-
Proptech
Automating Tax Lien Research
A tool that pulls tax-delinquent properties from county records, ranks them as tax lien purchases, and runs basic title searches.
-
Automation
Auto-Filling Agent Deal Submissions with Property Data
Rebuilding the agent partner portal raised agent engagement 35%, cut support tickets 40%, and increased repeat submissions 25%.
-
Self-Service
Borrower Self-Service at Equity Loop
A self-service portal for the appraisal step raised completed applications 20%, cut support inquiries 30%, and shortened time to funding 7%.