When AI agents do the work, trust becomes the design problem.
Chase Home Lending is rebuilding loan origination around AI agents to fund loans in days instead of weeks, but AI efficiency gains won't be realized unless our employees trust their work.
AI agents replace manual reconstruction with evidence-based review.
Agentic loan origination platforms are resetting the standard for speed and cost. After vendor solutions failed to deliver the expected efficiency, Chase chose to build rather than buy.
A loan origination system carries a mortgage from application to closing and stores its documents, data, decisions, and history. AI has expanded what Chase could build now that models can reliably read documents, check data, and enter information to move a loan forward.
I own the employee experience and UI engineering for the loan origination system, and lead two designers and an engineer on the project.
Mortgage competitiveness comes down to rates. Today, process and coordination inefficiencies force employees to spend most of their time verifying work instead of moving loans forward — and that overhead shows up directly in our rates.
The process scatters loan information across forms, documents, comments, and personal files. Every touch on a loan starts with piecing together what's already been done, which caps each person's capacity.
The system keeps loan data separate from the work.
The system logs what changed, never why. Employees have to reconstruct that context before they can act.
The workflow depends on employees to advance each loan, but the system doesn't show who's active, what's blocked, or what needs action. Handoffs go unnoticed, loans sit idle, and closing takes longer.
Work can wait without an owner.
Field observation showed how both failures shaped the workday.
Finding work and piecing together the loan history took longer than the decisions themselves.
46
Locating the next task
Employees monitored the pipeline because the system did not direct tasks to them.
Reconstructing the loan
Before acting, employees rebuilt the loan's history from comments, documents, and personal records.
The decisions only took minutes.
One processor spent ~90% of a two-hour session monitoring the pipeline. They could neither complete their only task nor schedule a follow-up in the origination system, so they copied the task into Salesforce.
One underwriter spent ~75% of a two-hour session documenting the loan in a personal spreadsheet, including custom income and debt calculations, before making the initial decision.
I designed the architecture to embed agent evidence throughout the application, so employees have the context they need wherever they pick up the work.
The new system replaces 100+ separately maintained forms with governed task types built from shared components. Four connected surfaces carry that evidence at every level of the work.
From finding the next task to verifying a single value
Shows which loans need attention and the latest agent activity on the rest.
Turns agent and human activity into a readable history.
Puts the evidence, rationale, and expected impact behind one required action.
Links verified information to its source and shows what still needs verification.
Pipeline: make waiting work visible
The current pipeline makes employees open a loan to see whether anything has changed. I added the latest agent activity, who is working on the loan, and any open tasks to each row. Loans that need a person appear in a separate section, so the queue directs attention instead of demanding constant monitoring.
Static loan rows become a visible work queue.
The design shifts responsibility for surfacing each handoff to the queue, instead of asking employees to poll for work.
Timeline: let the loan explain itself
The current system's comments are hard to follow. Showing every agent event would recreate the same problem at a larger scale. I kept each event as a structured record of who acted and what changed, then used AI to summarize older activity while keeping the full record one click away. Before-and-after values show how the loan changed over time.
Structured events preserve the facts while AI makes the history readable.
- Several earlier events, condensed.
The summary reduces reading without becoming the record. Employees can still inspect the event that changed the loan.
Task: put the evidence before the decision
Giving employees more evidence without structure would still leave them to decipher the loan. I standardized the task layout and limited the agent to supplying loan-specific content. Every task follows the same sequence: why a person is needed, what the agent used, how it reached its conclusion, and what the decision will change.
One task turns scattered context into a reviewable decision.
Unplanned work enters the same governed structure as agent-generated tasks, so it does not create a second, undocumented path through the loan.
Details: make every value answerable
I kept the complete field list out of the everyday workflow and reserved Details for verification. Employees can inspect every value, open its source, or filter to the information that has not been verified.
Verified information links every value to its source.
Verification becomes an exception path instead of the default way employees navigate the loan.
The system is still being built toward its first home equity release at the end of October 2026, so the honest measure is the record it will leave behind, not the screens. I own the employee experience and lead the UI engineering, and most of my time now goes to the trace schema: the contract for what agents report about their actions, and how that shows up wherever a person has to decide.
What I would do differently: define that schema before designing a single review surface. We designed the surfaces first and then discovered how much of their credibility depended on what the agents could actually say about their work. Cost per loan, cycle time, and loans per employee come after release; until then the work is judged by whether an underwriter can see why a value changed.
How do you get someone to trust work they didn't do themselves?
Employee Pattern LibraryEvery screen in this case is built from that system's components, down to the type treatment that distinguishes AI-generated text.