Doma set out to disrupt an industry running on 1890s procedures and 1990s technology, co-founded by Dr. Michael Housman, who served as Data Science Advisor on Doma Intelligence. The result: home closings in days, not months, and 84% of refinance orders underwritten instantly.
A centuries-old industry running 1890s procedures on 1990s technology,
gating the largest transaction of most Americans' lives.
The bet: build the new one AI-native.
Doma is a fintech company co-founded by Dr. Michael Housman, set up to revolutionize the centuries-old title and escrow industry. The traditional closing process combined 1890s procedures with 1990s technology: a workflow ripe for disruption, gating the largest transaction in most Americans' lives.
Doma's mission was to create a frictionless home closing experience using cutting-edge AI technology and high-quality service, from the company's inception. Not bolted on. Built in.
As a co-founder and Data Science Advisor, Dr. Michael Housman led the development of Doma's AI-driven approach.
The closing process was the cumulative outcome of a century of paperwork, multiple manual handoffs, and a customer journey nobody had ever designed. Four constraints defined it.
A multi-month wait between offer accepted and keys handed over, out of step with every other modern transaction.
Multiple manual procedural touchpoints between humans, systems, and counterparties. Every handoff a latency point.
Physical documents, scanned PDFs, signatures collected the way they were in 1985, dragging the whole system down to that speed.
The largest purchase of most people's lives, gated by the most opaque process they'll ever encounter.
Doma Intelligence is the AI platform built to address the industry's inefficiencies from the ground up: a comprehensive system of ML-driven underwriting, NLP for communications, computer vision for documents, and predictive risk models. Each component attacks a specific seam in the closing process. Together, they compress weeks into days.
100+ data sources fused across millions of transactions. The risk engine reads the deal in seconds, not days.
Trained on 100K+ emails. Routes, summarises, and replies to the high-volume back-and-forth that gates every closing.
OCR-grade document processing for the paper that title & escrow can't escape. Pixels into structured data.
80% of refinance orders underwritten instantly. Risk surfaced earlier, fewer surprises late in the process.
The most honest measure of an AI platform is the time it takes a file to move through it. Three and a half years after launch, processing time per order had fallen by 78% on Doma Intelligence, a step-change, not a trend line.
A better closing experience translated directly into wallet share: the percentage of an originator's pipeline routed through Doma. The platform became the default for partners who'd seen the difference.
Combined with improved customer satisfaction and clear industry leadership in AI, measurable business impact, not theatre.
Start with AI in mind: don't bolt it on after the workflow is set.
Combine sector expertise with data science: neither one alone is enough.
Build robust data infrastructure first: models live or die on the pipes underneath.
Foster cross-functional collaboration, between product, ops, legal, and ML.
Iterate continuously: the first model in production is the floor, not the ceiling.
Doma is what AI-native looks like when it's built in from day one. Most teams won't get to start from scratch, but the same components, applied to existing workflows, move the same numbers.
