Rocket Mortgage is quantifying AI returns by measuring the number of loans per team member. 

“If I think about what AI does best, it’s almost hard for me to think about a better case study than a mortgage company,” Brian Brown, chief financial officer and treasurer at Rocket Companies, parent of Rocket Mortgage, , said Dec. 3 during the UBS Global Technology and AI Conference 2025. 

Mortgage companies are human capital intensive and workflow intensive, he said.  

“It’s almost like a manufacturing plant, but rather than manufacturing a widget, you’re manufacturing data,” Brown said. 

For example, there are set limits and criteria in lending that AI can monitor for, LTVs and DTIs that will accepted, he said. By streamlining these lower lift efforts, employees can do more meaningful work. 

 “You don’t want a loan officer working a pipeline of low-intent clients and wasting time,” Brown said. 

When measuring AI returns, “it comes out in increased capacity, meaning you can do more loans per person.”  

Q3 AI results

Rocket Mortgage had more than 14,000 employees as of the third quarter, and its mortgage closed loan origination volume grew 13.7% year over year to $32.4 billion, according to its Q3 earnings presentation.  

Rocket also released its AI-powered Pipeline Manager Agent during the quarter, according to Rocket’s Oct. 30 8K filing with the SEC. The tool helps loan officers prioritize the right leads. Loan officers also use AI to create personalized messages to clients. 

In September, these tools resulted in a 9-percentage point increase in client follow-ups, according to the filing. 

With AI and automation, the mortgage company is saving 1.1 million hours annually, according to its Q3 earnings presentation. 

 Register here for early-bird pricing for the inaugural FinAi Banking Summit 2026, taking place March 2-3 in Denver. View the full event agenda here.