Previous AI in mortgage isn’t being held back by ambition; it’s being held back by infrastructure. Amy Gromowski, VP and Head of Data Science at Cotality, sits down with HousingWire’s Allison LaForgia to break down why becoming AI-First starts with one thing: data that’s ready to be used.
“AI has opened up a world of possibilities for everything we do,” Gromowski said. “There’s an individual personal productivity play. And then now we’re already in the era where there’s… a first billion-dollar revenue company with two people.”
That rapid expansion of possibilities, however, can create uncertainty about where to begin. “So I think that leaves us all feeling a little bit like, where do we start?” she said. “There’s a short-term play, and there’s a long-term play.” She continued, “If we just start focusing on the short-term… get familiar with the AI, and then move to enhancing and starting to streamline some of the existing work that you do.”
Once that initial foundation is in place, the conversation naturally shifts from experimentation to transformation. “We can reimagine mortgage. We can reimagine real estate,” she said. “But, I think in order to get to this real reimagined state, we need to take this in smaller increments . . . and once we have that . . . we can really start building towards a future state.”
A major shift enabling that future is the move from static dashboards to natural language interfaces. “The way that we have interacted with data and analytics has been either through UIs or APIs or bulk data transfers, but that’s very limiting,” Gromowski said. “With natural language, it completely opens up the realm of possibilities.”
That shift, while powerful, introduces new complexity behind the scenes. “When you’re a company like Cotality that has 16 petabytes of data, five and a half billion records . . . that’s a massive amount of data,” she said. “You have so many domains . . . hazard . . . geospatial aspects . . . I have to traverse all of that data and bring that back into a high-quality response.”
To address the challenge of understanding massive amounts of data, Cotality is investing in deeper infrastructure. “All the capabilities around interpreting the question with the semantic layer, building out knowledge graphs and how the data talks to each other and the relationships there and then, how to build queries at scale to traverse all of those different data assets,” she said.
At the same time, adoption depends on delivering that complexity in a way that feels simple to end users. “We are all about meeting the market where they are,” Gromowski said. “Whether that’s bulk data… cloud marketplaces… API… or UIs.”
The user-first approach is now evolving toward AI-native delivery. “We’re moving now to AI-ready data and MCP servers . . . companion files that you deliver along with bulk data that give the instructions . . . through natural language,” she explained. “Ultimately, this opens up the space for, you know, agent to agent.”
As capabilities expand, however, trust becomes the limiting factor. Gromowski highlighted how quickly sentiment can shift. “The question around trust in AI in the home buying process went from one in three people, so 33% said yeah, I would trust AI in the home buying process, to 16%, so it was cut in half in less than a year,” she noted.
That shift reflects growing awareness of AI’s power. “People are coming to the realization at scale here that AI is powerful . . . so what does that mean in the home buying process?” she said. “It’s a big financial transaction for the end consumer. It’s an emotional transaction . . . [the consumer] wants to know more about what that AI has advised.”
Because of that, transparency is no longer optional, but foundational to adoption. “It’s really important that the consumer is part of the journey, that we’re listening to the consumer and able to provide a level of transparency and quality to ensure that the adoption is there.”
Rather than replacing humans, AI should remove friction. “Let people do what they do best, which is engage with people,” Gromowski said. “Filling out paperwork, analyzing documents . . . that all puts friction points on people . . . if the AI just starts tackling administrative work, then that really frees up people . . . and builds trust.”
At the core of that vision is governance. “Responsible AI is king in my world,” she said. “Transparency in terms of being able to trace every step that an agent or AI is doing . . . what data is it tapping into . . . what’s the reasoning… and how it came to that.”
That includes rigorous oversight. “We have a whole AI Governance Committee . . . legal. . . compliance . . . monitoring . . . ensuring we’re compliant,” Gromowski explained. “Ensuring that every single one of your models are in a known state . . . that gives that stamp of trust.”
