Copperlane on the Fintech Hunting Podcast

We recently joined Michael Hammond on the Fintech Hunting Podcast to talk about what is changing in mortgage right now, and where AI is creating leverage for lenders.

Here are our highlights:

Mortgage doesn't need more software, it needs capacity

One idea we kept coming back to during the conversation:

"We are trying to build more of an employee rather than trying to build software."

Most mortgage technology over the past decade has focused on workflow tools. But the real bottleneck in origination is manual work required to move a file forward.

Loan officers still spend hours reviewing documents, chasing missing information, and answering repetitive borrower questions.

"When a borrower uploads documents, we are actively evaluating their profile. If we find anything that needs to be flagged, we can proactively reach out."

The shift toward agentic AI is about forward momentum. Instead of waiting for manual review cycles, files move continuously.

Mortgage is a context problem, not just a numbers problem

Mortgage workflows involve nuance. Income documents vary. Borrowers have non-linear financial histories. Information rarely arrives perfectly structured.

Earlier generations of AI struggled in these environments.

"People have been burned before. Earlier AI worked well for structured inputs, but mortgage is full of unstructured context."

Recent advances in generative AI allow systems to reason across documents, identify inconsistencies, and surface issues earlier in the process.

"With generative AI, there is more potential to understand people and context, not just structured data."

This is what makes AI useful in the middle of the loan, not just at the edges.

AI changes how lenders think about team structure

Mortgage has historically scaled headcount based on rate cycles. When volume increases, teams hire quickly. When volume drops, teams contract.

That creates operational volatility.

"If we can automate the middle of the loan work accurately, loan officers can focus more on relationships and keeping borrowers happy."

AI allows lenders to operate with leaner teams focused on borrower experience, while repetitive operational work happens in the background.

Instead of reacting to market cycles, teams can scale more consistently.

Trust comes from visibility, not black boxes

In a regulated industry, lenders need to understand how decisions are made.

AI systems need to show their work.

"Every decision our agents make is documented in a reasoning trace, so lenders can see why something was flagged."

Instead of adding more layers of manual review, AI can surface the highest-signal issues earlier in the process.

The goal is not to remove humans from the loop. It is to make the loop more efficient.

AI should work with existing systems

One of the biggest concerns lenders have is implementation risk. Replacing core infrastructure is rarely practical.

"We integrate with whatever LOS and POS a lender already uses. The AI runs continuously in the background."

Rather than forcing workflow changes, agent-based systems adapt to the environments they operate in.

Each lender has different priorities and structures. Flexibility matters.

Continuing the conversation

The lenders leaning into AI today are already seeing improvements in speed, responsiveness, and borrower experience.

The gap between early adopters and those waiting on the sidelines is starting to widen.

Mortgage does not need more tools. It needs leverage.

We appreciate Michael Hammond and the Fintech Hunting Podcast for the conversation.

Watch the full episode here

Listen to the podcast episode here