LoanWrite
LoanWrite is a recreation of an underwriting workspace originally built for another company. The original sits behind that company’s IP, so this clone shows the work without exposing it: every application arrives risk-scored, each file opens into a seven-tab review, and an AI copilot answers questions about the file — Regulation Z included.
- Client
- Recreation — original built for another company (their IP)
- Role
- Product design (recreation)
- Timeline
- Recreated as a clickable prototype
- Platform
- Desktop web app
Overview
Services: Product strategy, Workflow design, AI interaction design, Data-dense UI, Design system, Interactive prototype.
Underwriters spend their days switching between loan origination systems, document folders, credit reports, bank data and compliance checklists. The work is high-stakes and repetitive, and the context lives in too many places.
LoanWrite pulls it together. A single applications table with status, risk and ownership; a file view that walks from overview to final decision; queues for underwriting, identity, risk and document review; and an AI copilot that sits beside every file.
What if an underwriter could understand any loan file in two minutes — and ask it questions?
The problem
Triage takes too long
The problem: Every morning starts with a pile of applications and no quick way to see which ones need attention first.
What we did: One table with loan amount, status, risk score, submission date and owner. Risk is scored on arrival, so the riskiest files surface immediately.
For users: Underwriters start the day on the files that matter, not the ones on top of the pile.
Context is scattered
The problem: Personal details, loan terms, documents, fraud checks and bank data live in separate systems.
What we did: A seven-tab file — Overview, Personal, Loan, Documents, Fraud, Bank, Review — that opens on status, overall risk and processing time at a glance.
For users: Reviewers stay in one place, so nothing slips through the gaps between tools.
Risk is a black box
The problem: A single score hides why a file is risky, so reviewers re-derive it by hand.
What we did: Risk factor analysis scores credit history, debt-to-income, banking history, income and identity individually — with strengths and areas of attention called out.
For users: Reviewers can explain every decision — to the borrower and to compliance.
Answers require digging
The problem: Simple questions — the loan amount, the credit details, what Regulation Z requires — mean opening five screens.
What we did: An AI copilot in the sidebar with quick questions and free-form chat about the current file, plus built-in Regulation Z guidance.
For users: Answers in seconds, without leaving the file or reading the regulation.
Outcomes: One workspace, from queue to decision
- 7 tabs covering the complete loan file
- 4 dedicated workflow queues
- 1 AI copilot beside every application
LoanWrite is a clickable recreation, not a live product — every table, tab, queue and copilot interaction shown here works. The original belongs to another company and isn’t shown.