Grant Metcalf

Greater Chicago · UX and Design Leadership · AI Transformation

I build design functions, and bring the AI fluency leadership now needs.

Twenty years in UX and design, eight leading teams. I treat design as an operating discipline: the teams, standards, and practices that make good work repeatable at enterprise scale.

Contact me
2 → 12Team scaled from scratch, hiring every seat and building the operating model
100%Enterprise design system adoption across all product domains
3 TeamsShipping responsible AI tools built on standards my team set
90+Quality benchmark established for SME-graded AI research output

Selected work

01 Global Retail Pharmacy
Principal UX/UI Design Manager $6.4Mmodelled annual cost of having no system

Built the team, the operating model, and the design system that followed.

An enterprise pharmacy-software company had design happening in every product domain, but no shared discipline across them: separately built apps that did not feel like one product.

I scaled the team from 2 designers to 12, running every interview, hire, and onboarding myself, and put a senior tier in each domain with art-direction authority. Underneath that I built the operating model: shared ways of working, handoff templates, and a Figma component library built against Angular so nothing was lost between what design specified and what engineering shipped. I wrote the training curriculum and taught it myself, because a standard nobody is taught is a standard nobody follows.

The team was aligned. The company was not. Getting a company-wide design system approved took three pitches, because each audience heard something different: product teams heard speed, developers heard rework, executives heard cost. So I modeled what the company was actually losing to redlining and rebuilt components, just under $6.4 million a year, and that is what got the work resourced.

Separate apps became one system, adopted by every product team. The standards outlasted reorgs, turnover, and my own departure.

02 Enterprise Safety Science
User Experience Manager 3product teams building on the standards

Designed AI standards for people who have to sign their name to the result.

The people using these tools certify products. An internal AI lab had proven its models worked, but the builds validated the engine, not the workflow. No expert signs their name to an answer they cannot check. In certification that signature is the product, and its value comes from independent human judgment, not from the system that produced the answer.

That gap between a working model and a workflow an expert would sign off on was my team's to close. I led the UX portfolio across the lab's initiatives with the AI team and three product groups, and managed two designers embedded in those teams. Half the job was teaching those teams what UX could give them. The other half I did alongside my designers: the flows and entry points that placed each capability in a real workflow, and the MVP functionality that made it usable.

Research with the people who would use these tools pointed one way: design for verification instead of trust. My team set the responsible-AI standards behind every capability, so an expert can always check a result against its source. Three product teams build on them now, from a playbook we wrote so the approach repeats.

RESPONSIBLE-AI INTERACTION STANDARD What an expert sees before anything enters the record ENGINEERING TEST REPORT · DRAFT Test method summary AI GENERATED CONFIDENCE 87% SOURCE: TR-4821.PDF, P.12 Confirm and insert Edit Nothing reaches the certified report until the expert accepts it. 01 Output labeling Every field a model wrote is marked as such. 02 Confidence scoring The model's certainty is shown, never hidden behind a clean-looking answer. 03 Source traceability The original file and page sit one click away, at every confidence level. 04 Human confirmation The expert who signs the certification is the one who accepts the output. A high confidence score never removes the check. That is the whole design. Pattern shown as designed. Client interfaces are confidential.

When implementation later surfaced data and workflow problems, they were the ones the research had already flagged.

03 UX Research Function
User Experience Manager 90+first-pass quality bar, SME-graded

Stopped a UX research function from spending its AI time savings on cleanup.

AI was saving us time but then took it right back. Output arrived fast and rough, so every deliverable meant rounds of corrective feedback or rewriting it ourselves. No shared standard, no quality bar, and hard-won learnings vanished when the chat window closed.

I designed the approach and built it out: an agent that improves prompts by grading what they produce. Using thirteen quality dimensions, it scores raw output against a gold standard: UX-edited, then SME-approved. Where output falls short, the gap points to what the prompt is missing, so the prompt gets rewritten and run again. And again, until it clears 90 out of 100 on the first pass. The prompt is what improves, so the next person is not coaxing the model through five rounds of corrections.

The savings are real but not free. The team and I did the manual passes first, at full cost, so there was something to grade against. After that, processing discovery into a draft went from hours to minutes, and review from three hours to one, because output that clears 90 needs correcting, not rewriting.

EVALUATION LOOP How a prompt earns the right to be used by the team Skill file Complex instructions written once, versioned, run many times Raw model output First pass, before any human editing Graded 0 to 100 Thirteen dimensions, against a gold standard: UX-edited, then SME-approved Prompt revised The gap names what the prompt was missing BELOW 90 FIRST-PASS SCORE 100 0 90 FIRST-PASS BAR PROMPT ITERATIONS ADOPTED A prompt joins the team library only once it clears 90 unaided. That is what cuts the editing behind it. Model-graded against a UX-edited, SME-approved standard. A threshold, not a claim of absolute accuracy.

A quality threshold, not a claim of absolute accuracy.

In their words

Leaders and peers

“He exemplifies what you want in a UX leader. He delivers thoughtful, effective experiences and elevates the team around him. His ability to balance strategic thinking with hands-on execution makes him ideal for a senior leadership role in UX.”

Michael Henry · Senior Director, Transformation and Innovation

Grant's ability to visually communicate strategy and design intent is unmatched. He turns messy, multi-layered ideas into clean visuals, holding himself to the highest standard of quality to deliver polished, intentional work that drives real team confidence.

Andrew Margolis · Senior Product Manager

As a technical leader on the software team, what I valued was the usability, technical feasibility, and quality of his design work together. A natural leader, and a strong asset to any organization looking for a UX leader.

Phil Daniels · Enterprise Architect

He connected our UX journey maps, personas, and workstream requirements to change management without being asked. His curiosity and his use of AI produced insights our team could act on.

Lisa Becker · Senior Manager, Change Management

He took competing views on a new internal tool and built a proposal that created excitement and confidence around it. His approach shifted the team from apprehension to genuine enthusiasm.

Erik Leider · Conformity Assessment Program Principal

People I have led

“He embodies what exceptional leadership looks like. He provides clear direction without micromanaging, creates space for us to voice concerns, and advocates for our work-life balance.”

Sarah Lai · UX Designer

He takes full ownership and fosters a team environment of open communication. He proactively takes initiative in leadership, project planning, personnel hiring, and critical design decisions.

James Damonte · Senior UX Designer

He understands each designer's unique strengths and communicates in ways that connect personally. He spots where I can improve and turns it into an opportunity to grow. His feedback is specific and actionable.

Alex Molta · Associate UX Designer

He leads by example and supports others to lead the efforts they care about. His work ethic and communication created focus where there was often no direction across the wider organization.

Brian Fugate · Senior Visual Designer

He drives clarity and structure when the team faces ambiguous situations, and creates space for teammates to share ideas and improve how we work. He mentored me up as a designer.

Erik Gamradt · UX Designer

The offer

The enterprise work is covered by client confidentiality.

Rather than post sanitized fragments, I walk through the real artifacts live: the operating model, the design system, the roadmaps, the AI interaction standards.

Contact me

What UX is for

The view

Stories are the oldest technology we have. Before writing, before commerce, before anything anyone would call design, people sat around and told each other what happened.

Every cross-functional team holds frontline insight, from engineering constraints to what support hears all day. Listening is not a soft skill. Teams that listen get better. Teams that do not, plateau.

The same is true of an enterprise. Organizations rarely fail at design for lack of talent. They fail because what users struggle with never reaches the room where decisions get made.

Bridging that gap is the real job.

Screens, design systems, and operational frameworks are tools for carrying human reality into business strategy. Build that bridge well and design stops being a step in the delivery process. It becomes the foundation of how enterprise decisions get made.

How design scales

01

Culture follows what is visible, not what is announced

Announcing standards rarely changes behavior. Showing the work does. I run practices like Walk the Wall, putting active work into shared cross-functional space on a predictable schedule. Once quality moves out of private critique and in front of product and engineering, the standards enforce themselves.

02

Interoperability relies on shared language, not shared taste

Scaling design across enterprise domains takes more than creative consensus. It takes a shared taxonomy, common vocabulary, and clear system governance. Unglamorous work, and the highest-leverage kind: it removes friction, aligns multidisciplinary teams, and frees designers to solve business problems instead of renegotiating basics.

03

Capacity planning builds cross-functional trust

Design loses credibility when its commitments lack rigor. I estimate design against the same point scale engineering uses, which makes our bandwidth legible and defensible in sprint planning. Research, cross-domain alignment, and rework are priced in, not absorbed. Stakeholders get what was promised, on the date promised.

04

Continuous research drives predictable roadmaps

Research treated as a kickoff formality wastes budget and misaligns teams. I embed research cadences into sprint cycles so insight arrives continuously rather than once. We weight toward high-friction workflows and time-constrained users, where the workarounds live. Metrics show where people stall. Research shows why. Neither is trustworthy alone.

Growing people

01

Effective management requires situational leadership

High-performing teams are not built on one-size-fits-all management. Different designers need different amounts of autonomy, constraint, and direction depending on the work and where they are in their career. So I read the person, not the org chart. Tight guardrails for some, wide runway for others. It maximizes output and builds psychological safety at the same time.

02

Retention is built into organizational architecture

Top talent stays when career growth is clear, intentional, and structural. I build it into the team through transparent competency frameworks, clear leveling paths, and senior tiers with real mentorship expectations. Then I align high-leverage business work with what people want to become. Engagement rises, costly turnover falls, and the biggest career leaps happen here.

The detail

Looking for the specifics? The resume has them.

Roles, dates, and the twenty years that did not fit into three case studies.

Read the resume

About

Grant Metcalf

Hi, I'm Grant. I build design functions and the teams that run them.

When I'm not running teams, I'm a devoted father. I draw and I read. I love storytelling, problem solving, and learning new things, and I travel whenever I can. I love learning the history and culture of new places.

As a people leader, I would rather build the system than stand over the work. Give people a clear destination, access to the right tools, and enough room to fail fast, and they will hand you back something better than you could have imagined.

Chicago based. Looking for a Director role where I can write my next chapter.

Next

Let's talk about the work.

The fastest way to see what I have built is to have me walk you through it. Send me a note and I will reply the same day.

Contact me
Grant Metcalf · UX and Design Leadership LinkedIn Resume Email