Enter the password to continue
.about
Most of my projects start the same way: no documentation, no UX history, and a dev team already months ahead. I sit with users, figure out how the work actually gets done, and ship quick fixes while the bigger redesign comes together.
I've spent 16 years in FinTech, insurance and other regulated industries. I've led teams and hired designers, and I still build components myself. Lately I use AI to cut the grunt work so more of my time goes to the hard problems.
.contact
.case-study
MetLife moved its design system from XD into four new Figma libraries covering every global market, and updated it for an upcoming rebrand along the way.
Impact
100% migrated to Figma. Every enterprise web and mobile product now builds from the new libraries.
Team
I was one of two design leads in a pod with a Scrum Master, a Product Owner, a Product Manager, three designers (plus whoever we could borrow) and several developers. We ran two-week sprints in Azure DevOps, with daily standups, demos every other week and quarterly PI planning.
The lead role was hands-on. I built components, reviewed every designer's work before it shipped, QA'd every build, and merged and published all four libraries.
File Structure
The system lived in five Figma files. One token file fed color, spacing and type into four libraries: UI, Brand, Conversational AI and Responsive Web. Each sprint, a relay file handled dev handoff and a templated deck covered the stakeholder demo.
Requirements
Along with the move to Figma, we improved how components worked and updated the visuals ahead of a rebrand.
Problem
Multi-step forms, permission-based views and edge cases don't fit Figma's prototyping tools. Developers were left guessing from static screens and meeting notes.
Solution
I added flow diagrams and walked developers through every conditional path before each release, so the build matched the design the first time.
Changes Needed
Designers new to Figma or MetLife also needed documentation they could actually follow.
Problem
Redlines can't show timing or easing. Even 4 to 6 animation options took days to make, so refinement stalled before it started.
Solution
I paired Adobe Animate and Jitter with Figma Make to generate 10 to 20 options up front, then added a motion review with developers before build. First drafts went from days to hours.
Solution
Variables had just launched. I used them for all color, which gave us tighter control and fewer component variants.
Documentation lives in the library and inside each component, so it's there when people need it.
Problem
Many MetLife publishers use Adobe Experience Manager (AEM), which strips out a lot of what Figma components can do, especially nested components.
Solution
I ran walkthroughs with publishers and AEM owners until I knew exactly what AEM could handle, then designed around those limits.
Governance
Live libraries drift. Names slip, components get orphaned and variants go missing, and someone has to catch it before it reaches 25+ teams.
Problem
A manual audit meant inventorying every component, checking each variant, logging issues and writing a report. It took about two weeks per pass, and MetLife blocks Claude and most AI tools, so there was no shortcut.
Solution
I got an exception approved to connect Claude to the live Figma files through the Figma MCP. It walks all four libraries and flags naming drift, orphans and missing variants. A designer reviews every finding before anything changes. The audit now takes about 15 minutes.
Same audit, two ways. By hand it took about two weeks. With Claude and a designer checking every result, it takes about 15 minutes.
Accessibility
When a component changed a lot in function or look, I built a simplified version in a separate library instead of changing the core one.
That kept existing work from breaking and gave non-designers something easier to use.
.case-study
Turning MetLife's 20-year-old plan setup and management systems into one role-based app.
Impact
Up to 80% of MetLife employees use these systems. They now work in one app instead of several.
The Challenge
In early 2024, UX was brought into a modernization project for MetLife's plan setup and management systems. Development was already well underway, with no designers involved.
The goal was to fix this system and set a better modernization process for every MetLife enterprise app after it.
Problem
The project kicked off with no UX at all. We joined four months in, and proper discovery took three more, which put design 6+ months behind development.
Solution
I split the work into parallel tracks. First, I audited what had already been built and shipped quick fixes: font stacks, spacing, layouts, page templates and a full accessibility audit. At the same time, I designed MVP screens for developers and a longer-term prototype with a new Enterprise design system.
Project Timeline
Development had been running for over three months before UX joined. To catch up without stalling delivery, quick tactical fixes ran alongside full discovery and modernization.
Problem
MetLife's existing library was a loose set of marketing components, not built for an enterprise app.
Solution
During the three-month discovery phase, I also built a new Enterprise design system. Its 37 components became the foundation of MetLife's Enterprise UI library.
Research
Four main roles, each with different needs and experience levels:
User Pain Points
Current State
Users logged in and out of several systems all day, usually with two or more sources open side by side.
Problem
The app had nearly 10,000 input questions, many for products that no longer exist. Most had help text, and the tone and style were all over the place.
Solution
I brought in a content writer and ran a content audit with SMEs and product owners to find the dead questions. Developers queried which ones were rarely or never answered. We delivered a style template so product owners could keep the content consistent.
Solution
The goal: bring every system and outside reference into one web app.
Nobody has to switch between tools or enter the same data twice.
Solution
Each role now sees only the parts of the app it needs.
Navigation got much simpler, and the path to daily tasks is clear.
There's less to wade through, and new people learn the system faster.
Problem
The system spans 60+ roles. In discovery, the biggest complaints were getting lost and information overload. Most people only used a small slice of the app.
Solution
I focused on 4 core and 6 secondary users and made the app role-based, so each person only sees the features and data their job needs.
Global Impact
37 new components, which became the foundation of the Enterprise design system now used across MetLife's apps.
Results
Users who used to juggle [4] systems now work in one. Setting up a new plan went from [5 days] to [2], and new hires ramp up in [6 weeks] instead of [4 months].
For thousands of MetLife employees, the workday now runs through one app built around their role.
.case-study
Replacing a set of disconnected legacy tools with one portal for Freedom Mortgage employees and broker partners.
Employees and partners worked across a string of disconnected legacy systems that were dated in both function and looks.
I was the only UX/UI designer. I ran every UX and creative workshop and produced and presented all deliverables.
Overview
Employees and partners moved between legacy tools all day with no single source of truth. That meant wasted time, errors and frustration at every level. The goal was one modern portal built around how people actually work.
Problems
Problem
No documentation and no power users meant years of workflows and business rules lived only in people's heads.
Solution
I went to the New Jersey offices and sat with people in different roles while they worked. That's where the undocumented steps came out, along with problems people had stopped noticing because it was just "the way things work."
Strategy
Research
User research, competitive analysis and information architecture to understand the problem and the people.
Problem
So many user types needed a voice in discovery that in-person workshops weren't practical.
Solution
I ran them remotely, using UserTesting.com for A/B tests and prototype feedback and Optimal Sort for card sorting. I added an open feedback field to the card sort, and it turned out to be one of the best sources of feedback in the project.
Design
Wireframes, high-fidelity designs and interactive prototypes, tested with stakeholders and users along the way.
Problem
The first few sprints meant a lot of rework. Developers raised concerns at the end-of-sprint demo, after the design work was already done.
Solution
I added a developer-only review two days before each stakeholder demo. Technical issues came up early, I fixed them before the full review, and rework dropped.
Key Feature
The biggest new feature was a pricing tool for brokers and partners.
It replaced manual calculations, which cut errors and sped up pricing.
Problem
The full project ran 2+ years, but Freedom wanted the pricing tool out right away. It was brand new and didn't belong to any existing system.
Solution
I built it as a standalone app that launches from BPM, their main broker portal, with clear messaging about the handoff. When an estimate is done, users can go back to BPM or keep working.
Visual Design
The new portal kept the Freedom Mortgage brand but brought it up to date.
Consistent patterns, better typography and more breathing room made it feel more trustworthy.
Results
The new pricing tool cut loan estimate time from [20 minutes] to [3]. Partners went from logging into [4] systems to one portal, and pricing errors dropped [40%] in the first [quarter].
.case-study
Business software for small and mid-size construction companies.
Overview
Contractors run bids, budgets, contracts, time tracking, scheduling and reporting from any device, all synced with QuickBooks.
The Project
One place for contractors to manage their projects, on whatever device they have with them.
My Role
I was the only UX/UI designer. I ran every UX and creative workshop and produced and presented all deliverables.
For the first two months, I was also the project manager and business analyst, working with the two founders on requirements and the roadmap.
Pre-Discovery
Build a multi-tenant product that helps business owners actually use their business data.
Discovery
Define & Design
Retrospective
It was 18+ hard months at startup pace, and it worked. This was my first product from start to finish, and it taught me the most about the strategy side.
Knowify went on to lead its market and partnered with Intuit QuickBooks in 2021.
.case-study
A new intranet for a national law firm.
Overview
Nelson Mullins Riley & Scarborough LLP hired me to design the firm's new intranet, from wireframes through build.
My Role
I led every phase of UX and UI with the firm's stakeholders:
Process
Getting the requirements right and turning them into high-fidelity wireframes, with three rounds of stakeholder review.
Process
Once wireframes were approved, I designed the visuals to match the firm's brand, with five rounds of review.
Deliverables
The final design made the intranet easier to navigate and content easier to find.
One set of patterns kept every page consistent while still handling very different kinds of content.
Results
I wrote a style guide so the build, and every update after it, stayed consistent.
I worked with the developers through the whole build, so what shipped matched the design.
.case-study
Navian Capital launched in 2018, incubated by Bank of America Merrill Lynch and Morgan Stanley. It runs an independent platform for structured products and annuities.
A UX audit and full redesign of its Luma platform.
Overview
Navian's tools started as a sprawling proof of concept. They needed to become one platform.
The Project
I worked on both design and strategy across the whole effort:
My Role
Research
I interviewed users in several ways to understand how they worked and where the platform got in the way.
What I learned there drove the redesign.
Discovery
Stakeholder sessions and user research showed where the platform needed the most work.
Design
Working with Navian's developers, I redesigned Luma to fix the usability problems we found.
I simplified navigation and the hardest workflows and updated the visuals, while keeping it familiar for existing users.
Results
One platform that supports every part of Navian's business, from training and product creation to sales and support.