My Role
Title
Lead Product Designer at a leading California research university
Timeline
4 months (April 2025 - August 2025)
Tools
Figma, Figjam, Gemini
Key Outcomes
I inherited hundreds of pages of dense requirements with stakeholders stuck debating fonts over workflows. Through workshops and structured checkpoints, I accelerated approvals.
User interviews with staff revealed how they actually worked, not how the requirements assumed they did, surfacing critical student data needs and eliminating unnecessary workflow steps.
Where design had previously failed, I had to rebuild credibility from scratch. By resetting expectations and delivering efficiency gains, I established a new model for user-centered design.
Context & Challenge
I inherited a project where design reviews had stalled for months due to reactive reviews and unclear authority.
At a leading California research university, the Graduate Studies division relied on a decades-old desktop application to manage student records. Although still in daily use, it was outdated and increasingly brittle.
The new graduate management platform was envisioned as a transformational replacement: handling everything from petitions to degree conferral across ten major feature areas and hundreds of pages of FERPA-compliant requirements.
The previous designer had approached this like a visual design problem, asking stakeholders about button preferences and letting them art-direct font sizes in meetings. Review sessions had become entirely reactive, and resulted in months of stalled progress.
Design researcher Sarah Gibbons from Nielsen Norman Group pinpoints exactly why this approach fails. “In order to analyze a design and whether it meets its goals, there must be agreement on the problem that needs to be solved. This likely means a clear understanding of users and their needs. Without these, any feedback is subjective and baseless.”
The organizational dynamics compounded the challenge. Without user research or planned workflows, business analysts had naturally stepped into design decisions in the absence of UX expertise.
My arrival shifted established power dynamics. I had to prove that structured, user-centered design could break this cycle and deliver a system that actually reflected how people worked.
Discovery & Strategy
I established new expectations for collaboration, then discovered how staff really worked through user interviews.
I knew I couldn't just jump into new designs. I had to rebuild trust in the design process itself. The key was my kickoff presentation where I explained what UX was (and wasn't) and why we were taking this stepwise approach.
I needed everyone to understand that functional requirements weren't the same as user experience, and that our collaboration would look different from what they'd experienced before. That reset expectations and gave us a foundation to build on.
So I started where the previous designer hadn't: talking to the staff who would actually use the application.
I discovered that in the legacy software they had developed workarounds where they used several applications across multiple external screens. The requirements documents didn't capture this nuance or the inefficiency of their current process.
The user interviews revealed how they really worked. Users needed certain information to be persistent throughout their workflow: student ID, GPA, and other key data points that they referenced constantly while conducting business in the application.
This insight led to foundational design solutions: student search as a global starting point, and a static student header with key information that persisted across the application.
Design Process
I used workshops to build consensus and shift stakeholder feedback from visual debates to user-focused problem-solving.
Armed with those insights, I ran information architecture workshops over five sessions with business analysts and Graduate Studies division leads.
After poring over hundreds of pages of requirements, I led the first workshop with a summary of learnings from user interviews to ground the session in user needs, then we collaboratively grouped and prioritized data fields together in Figjam.
Those IA workshops became the foundation for low-fidelity wireframes that mapped requirements into actual workflows. When I walked stakeholders through these wireframes, something shifted.
Users had a hand in how information was organized, and I could advocate for the persistent data needs that they had described. For the first time, everyone was solving the right problem, as a team.
Stakeholders started engaging with the design decisions rather than debating visual preferences. For instance, instead of debating button colors, we were discussing whether the petition approval flow actually made sense for the user. This is exactly what I had hoped the workshops would achieve.
Only after reviewing, debating and approving these workflows in low-fidelity did I move to high-fidelity prototypes. Each stage worked as a checkpoint: IA structure first, then workflow, then visual design.
Design progress finally had a path forward. For stakeholders exhausted by months of chaotic reviews, this process was transparent, inclusive and decisive.
Before
Before I joined, reviews fixated on visual details, and without understanding user workflows. Approvals spiraled and progress stalled.
After
Designed with user research and UX best practices, not guesswork. The application keeps important fields persistent and organizes fields exactly how staff said they needed it.
Outcomes & Impact
What started as a fragile, mistrusted process became a repeatable model for how UX could succeed in the university’s academic systems.
From critique to collaboration
After business analysts raised concerns about deltas we had previously covered after handoff to dev, I pivoted by introducing delta walkthroughs early on in the following epic and explained where designs intentionally diverged from requirements and why.
Then, I proposed a structured validation checkpoint where business analysts could review designs against the FRDs earlier in the process. I encouraged async commenting in Figma, giving business analysts the space to evaluate designs outside of live meetings.
Process Impact
Approvals that had previously stalled for months cleared in a single cycle, with workshops eliminating reactive design feedback from the process. The guided approach transformed stakeholder dynamics from reactive critiques into problem-solving sessions.
Leaders appreciated the efficiency of the UX design process compared to prior chaos, and the approach proved scalable across the project’s many epics, becoming a repeatable model for success inapplication design.
Leadership Reflections
Success required equal parts design craft and stakeholder alignment.
UX leadership in complex organizations is fundamentally about alignment. Aligning requirements with workflows, stakeholders with process, anddesign with trust.
Execution meant creating a process that didn't exhaust everyone involved, where approvals moved faster instead of stalling indefinitely.
Collaboration meant earning trust from stakeholders who'd been burned before, bringing skeptical business analysts along, and working with transparency.
Once we achieved that alignment, we didn't just move the project forward; we created a replicable model for user-centered design in a requirements-heavy organization.
How I Used AI to Accelerate UX Workflows
I used AI as a thought partner and efficiency multiplier, not a replacement for design thinking.
Working as the sole designer on a project this complex, I found myself turning to AI tools for support in ways I hadn't expected. Large language models couldn't parse the dense FRDs for me, but they became surprisingly valuable collaborators in areas where I needed speed and a fresh perspective.
Thought partnership
Being the only product designer meant no one was challenging my assumptions or catching my blind spots. I created "UX Director GPT"; this AI became my thought partner, helping me stress-test user flows and surface issues I might have missed.
Accessibility
I used GPT to help design keyboard navigation across the complex application. I'd never worked with keyboard navigation before, and having an AI guide me through best practices saved hours of research while ensuring I got it right.
Error-state coverage
AI helped out with the tedious but critical work of defining error states and error message copy. Instead of brainstorming every validation scenario, I was able to focus on the big picture UX decisions while ensuring coverage of all the details.