Case Study 02 · Real-Time Inquiry Portal · Onity Group

I designed a real-time portal that replaces email threads with live case visibility.

A shared, live record for every loan inquiry, where mortgage clients can track progress, message teams inline, and skip the bi-weekly status meeting. Built to stop the "where is my case?" emails and hand analysts twelve hours back every week; prototype testing backs both.

Launched · 2026 Enterprise Software Interaction Design Mortgage Servicing Research-led
Real-Time Inquiry Portal dashboard, a central hub showing every loan inquiry in one organized, live view
A central hub, every loan inquiry in one organized, live view
62%
Faster case resolution
early data, first cohort vs. the prior email baseline
6 hrs
Weekly hours back per account manager
early data; AMs had spent ~12 hrs/week chasing status by hand
0
Status meetings by design
the live dashboard replaces the bi-weekly call

Role

Lead Product Designerend-to-end research, IA, interaction, visual

Team

Onity Group1 PM, 2 engineers, 1 designer (me)

Timeline

5 monthsdiscovery through launch, 2026

Tools

Figma · Figma Make · Claude Design · Copilotdesign, prototyping, and AI-assisted build

A note on confidentiality

To respect Onity Group's confidentiality, I've changed the visual treatment, screen names, and workflow details shown here. I chose not to password-gate this case study, since gates only add friction for the people I want to reach; instead I recreated the screens to keep the work open and compliant. The approach, decisions, and outcomes are intact and mine. For the real product and a deeper walkthrough, reach out.

Snapshot

The short version.

If you only read three paragraphs, read these. Problem, change, and results in about a minute.

Problem

Mortgage clients had no way to track loan inquiries after submission. Updates lived in scattered email threads, and account managers burned roughly 12 hours a week chasing status by hand and running bi-weekly meetings just to stay in sync.

Change

I designed one shared workspace where every inquiry has a live record: status, threaded conversation, attached documents, and a clear resolution path. Account managers oversee their full portfolio from a filterable table; clients see exactly where each case sits.

Results

Before launch, 90% of users located case status unaided in testing and no conversations were lost. Now live with a first cohort of three client orgs, early data shows about 6 hours back per account manager each week, against a baseline of roughly 12 hours spent chasing status by hand.

Problem area

Lost in the back-and-forth.

Once an inquiry was sent, there was no shared place to see progress, share documents, or message the team. Updates lived in hundreds of email threads, and account managers burned about 12 hours a week updating spreadsheets and running bi-weekly status calls to stay in sync.

Zero visibility after creation

Once a case opens, clients can't see status or owner. So they email "just checking in."

Fragmented messaging and files

Loan documents and replies scatter across email threads. Latest version, last word, both hard to find.

Meeting-heavy status management

Account managers update spreadsheets by hand and run bi-weekly status calls to stay in sync.

"Give me the visibility to see where we sit. Show me if I'm 20% or 80% done, simple milestones or a small chart."
Ivan Camarena · Client User
"I need to see all of our open cases, not just mine, so I can cover when someone's out."
Glenda Brue · Client Administrator
Solution

A central hub for real-time oversight.

One place where every loan inquiry lives, status, owner, key dates, updated in real time. The "where is my case?" emails stop, and clients can finally trust the system as the source of truth.

Dashboard view

One screen, every case at a glance.

The entry point. Status summary cards surface aging cases at the top; the live, filterable case table replaces the bi-weekly status meeting.

"I need to see all of our open cases, not just mine, so I can cover when someone's out."
Glenda Brue · Client Administrator
1

Every case state, the second you log in

Aging cases surface at the top, so clients see progress at a glance instead of emailing "is this done yet?"

Visual Hierarchy · Fast Feedback
Dashboard status summary cards row showing Open Inquiries 4, With Onity 2, Waiting on you 2, and Total Cases 6 counts
2

One table replaces the status meeting

A filterable source of truth for every inquiry, so most of the 12 hours a week of spreadsheet updates and status calls goes away, about 6 hours back per account manager.

Data Transparency · Recognition
Live, filterable case table with inquiry ID, subject, status chips, last update, loan number, owner, created by, create date, and days columns
Case detail view

Detailed insights and interaction.

A full-page case view that puts progress, conversation, and documents in one shared history, so clients and account managers stop hunting through email threads for the latest state.

"Every case is a different email thread, finding the latest update means hoping it didn't get buried."
Synthesis · across client interviews
1

Progress you can read at a glance

Clients see whether a case is starting or nearly done, so they stop emailing to ask, and know when it's their turn to act.

System Status · Visual Cues
Real-time progress tracker bar showing four stages: Submitted, Assigned To, With Onity, Resolved, each with date stamps, with tabs underneath for Overview, Conversation, Documents, and Resolution and Escalation
2

Messages and files, kept with the case

Chat and documents live in one shared history, so nothing gets buried in an email chain and a teammate can step in cold.

Information Access · Context
3

A clear way to close, or escalate

Clients confirm a fix or escalate to a manager, and a rating tells the team whether the outcome landed.

User Control · Feedback Loops
Resolution and Escalation tab showing Resolution Details with an Outcome dropdown, date field, closing comment, and Escalate plus Confirm and Close action buttons, next to an About Escalation help panel
Final Design

The full picture, tab by tab.

Four states of the expanded case-detail flow as a client moves through it: opening the case to see the overview, reading the conversation, attaching documents, and finally resolving or escalating. The dashboard chrome stays constant; only the inner panel changes with the active tab.

Validation

Did it work?

The portal launched to a first cohort of three client orgs, starting with the most-used tracking moves before the heavier tools land in later releases. Before launch I pressure-tested the direction with clickable-prototype usability tests; since launch, early data from real account managers backs up what the testing predicted.

90%

Success in status tracking

During testing, almost every user was able to find the exact stage of their case on the first try without any guidance.

Zero

"Lost" conversations

Users reported that having the chat and files tied directly to the case made them feel more confident about the history of the request.

6 hrs

Back per account manager each week

Early data from the first cohort: the portal's live record replaced the manual status reports account managers had been rebuilding by hand.

Post-launch

What I'm watching now it's live.

Now the first cohort is live, I'm tracking these three areas to confirm the portal is doing the job. Each one ties back to a concrete change in behavior, not a vanity number.

Shift in communication

I'm watching whether "status update" emails fall as portal logins rise, the signal that users trust the dashboard for answers.

Document task completion

I'm tracking how quickly clients upload requested files through the portal versus the old email-in method.

Self-serve reporting success

I'm measuring how often the "Export" and "Filter" features get used, a proxy for account managers handling their full workload self-serve.

Key Decision

Why I built a shared record instead of notifications.

The original ask from stakeholders was simple: add email notifications so account managers know when an inquiry status changes. I pushed back. Here's why.

Round-one lo-fi wireframes comparing the rejected notifications-only direction against the shared case record that was built, with annotations on why notifications leave the inbox as the source of truth
The round-1 lo-fi that settled it: notifications only vs. a shared record. Recreated; originals under NDA.
Why a shared record instead

Notifications tell you something happened. A record tells you what happened.

Email notifications would've reduced some back-and-forth, but they would've left the source of truth scattered across inboxes. Two people can get the same notification and still have different versions of what's true. The shared case record meant one party couldn't operate from a stale snapshot. Every status change, document upload, and message sits in the same place, so "what's the status?" becomes a lookup, not a conversation.

What I traded away

It was a harder build, and I had to make that case to engineering.

Notifications would've shipped in two weeks. A shared real-time record took three months. I framed the tradeoff this way: notifications reduce the number of follow-up emails; a shared record eliminates the reason for them. The 62% resolution improvement we modeled from prototype task times against the email baseline backed the call, but the harder moment was convincing the PM that the larger investment was worth it before we had any data. I used the research to make the case: six separate interviewees described the same "I sent an email, now I wait" pattern, and that specificity mattered more than a prototype.

Retrospective

What I took from this.

Two layers of impact that don't show up in the metrics. The work that shaped how I designed it, and the patterns the team kept using after I shipped.

Impact on me

Designing a system of record means defending what doesn't get built.

The first round of stakeholder asks would have stuffed every report, alert, and integration into one screen. Holding the MVP to four jobs, status, conversation, documents, resolution, took more meetings than the design did. It's the reason the prototype tested cleanly on the first pass.

  • 10 user interviews shaped the four-tab IA
  • 3 prototype rounds before usability testing
  • What didn't work: the SLA and escalation logic was designed without engineering in the room. It surfaced late, broke in review, and forced a re-spec before round three
  • What I'd do differently: bring engineering into wireframe reviews two weeks earlier, so that failure happens on paper instead
Impact on process

The patterns left behind for the next portal.

This was the first case-management surface on Onity's mortgage platform, so the components and the research artifacts both became templates. The next team building a similar workspace started from the moves I made here, not from scratch.

  • Live case-table pattern, status chips, owner, aging, added to the internal design system
  • Progress-tracker component reused on the Loan Deboarding portal
  • Interview synthesis template now used across the platform team