Design Leader  ·  UX and Product  ·  Physical and Digital  ·  Boston, MA

Patrick Lukasak

I take ambiguous, first to market opportunities and turn them into products people trust, and build the teams that ship them.

My depth: trust between people and products, in places where a mistake can hurt someone. Operating rooms, factory floors, vehicles, and wherever you exercise.

$0MyrProduct category created 0 to 1. With its CrossFit line, about 27% of Reebok US revenue
0Patents: 8 granted, 6 pending
0Across medical, industrial, auto, consumer

4 design teams led

0Adidas: Global Design HeadID / Graphics / Material / Creative Direction
2 → 9Medtronic: Senior PrincipalUI / UX / Research / Design System
0PTC: UX StrategistStrategy / UI / UX / Research
0Immuta: Design ManagerUI / UX / Research / Design System

Selected Work

Four problems that had no playbook.

Medtronic · Surgical Robotics In development

Luna

Getting a surgeon to trust a robot navigating a place neither of them can see.

Luna is Medtronic's first to market lung health robot. The hardware was already ahead of the market. The real problem was trust: reaching a lesion means threading a dense, branching maze of airways, and the clinician has to know exactly where they are, where the target is, and how to get there without losing orientation. If they cannot trust what the system shows them, the accuracy is irrelevant. I owned the software experience from 0 to 1 and built a live augmented pathway over the bronchoscopy video, with a navigable 3D model the clinician can rehearse before committing. The system shows the right data at the right moment. The human always stays the expert.

0 to 1, first to market Live AR navigation 12+ rounds of formative research Multi persona OR workflows
The call
Framed the problem as trust, not features. Established and scaled the Rapid Research program, a fast research cadence with a standing surgeon panel that now generates requirements shaping the product strategy.
AI
AI is integral to the Luna software, and the clinician always keeps the final say. I also built two AI agents to help run the research.
Team
Partnered with industrial design, human factors, engineering, and research to deliver one coherent system.
Role
Principal UX. Owned the software design end to end.
Navigation screens, plus four pages from the Luna style guide. Select one to expand it.
Medtronic · Surgical Robotics Shipped Patented Directed 4 designers

Hugo, alert fatigue

Surgeons were dismissing critical alarms the instant they appeared.

The alarms were written in engineering language, and if the robot let a surgeon keep operating, they would, whatever the alarm warned. I shadowed live and mock procedures, audited all 265 system notifications, and re categorized them around the surgeon's real time decisions instead of the system's internal wiring. Then I rebuilt alarm priority around the robot's actual recovery path. I got Clinical Safety and engineering to share ownership of the outcome, and started with the highest frequency alarms so the team could see the win at zero risk. The work became the reference model for how notifications are designed across other Medtronic OR products.

Notification Stops 265 → 32 High priority causes 104 → 29 SUS 54 → 69 Two patents
The call
Asked for no rebuild up front. Started with the 30 most triggered alerts, proved the win at zero risk, then earned the full system change.
Team
Clinical Safety, systems engineering, and the roadmap manager I partnered with to fit the work into the plan.
Role
Led UX and research strategy end to end, from field research to rollout.
265 notifications shipped in Hugo RAS v1
Notifications 265 → 32 High priority 104 → 29
104 high priority causes after the review
Hugo alarm relevelingFlow of 595 alarm causes from legacy priority to new priority after a decision tree review. Legacy high priority causes fell from 104 to 29: 43 moved to medium, 20 to low, 10 to informational, and 2 were archived. Legacy medium, low, and informational causes kept their level.Legacy priorityAfter reviewHigh104Medium256Low56Informational179High29Medium299Low76Informational189Archive2
595 alarm causes reviewed with a decision tree and interlock rules
View as table
FromToCauses
Legacy highHigh29
Legacy highMedium43
Legacy highLow20
Legacy highInformational10
Legacy highArchive2
Legacy mediumMedium256
Legacy lowLow56
Legacy informationalInformational179
PTC · Industrial AI Shipped · Directed 4 designers

Digital Performance Management

A factory floor that could measure its losses but never locate them.

Standard metrics tell a plant it is losing productivity, not where. In a plant of a hundred machines, maybe ten actually throttle throughput, and teams pour effort into the other ninety. I designed one closed loop: see the loss where it happens, locate the real constraint, act on it, and verify the data confirms it is gone. I built the operator dashboard and the action tracker myself, concept through handoff, and set the direction for the rest of the modules, directing four UI designers.

Closed loop: see, locate, act, verify Directed 4 UI designers PTC reports Thermo Fisher recovered double digit OEE
The call
Made seeing and acting one loop instead of a report. Then cut the richer charts when 4 of 5 plant managers could not read them, and went simpler than felt comfortable.
Team
Directed four UI designers, with live plant access through a joint development partnership with Rockwell Automation.
Role
UX strategist. Created the research strategy for the project, built the operator dashboard and action tracker myself, and set direction for the rest.
Three bottleneck views, the operator dashboard, and the action tracker that closes the loop. Select a screen to expand it.

Also in the portfolio

PTC · AIMVP built

PTC Orbit

An ML first view of an entire product fleet, unifying data from five enterprise systems. A core team of two: me and a Director of Product Management. I scoped the problem in 2018, years before the industry built products around it, developed the strategy and led 2 UI designers to complete the concept.

Immuta · Big dataShippedManaged UI designer and 2 front end developers

Immuta data access

Redesigned the core product feature experience: writing complex data access policies. 40% fewer user errors and a 55% gain in interface accessibility. Helped build the company's first design system, and research process.

Systems I built

Tools that raise the whole team and scale design impact across the organization, not just one screen.

Medtronic Surgical · Built with AI

OSDS design system

The design system behind all of Medtronic's surgical products, $6.5 billion worth. I created and manage it with Claude and Figma. The rest of Medtronic, a $135 billion company, is now adopting it organically.

Medtronic · Luna · AI

Agentic AI: 2

Built two AI agents to help the team analyze research, identify patterns, and provide design decision logic.

Built an agent to create, modify, and update our design system, including guidance, do's and don'ts, and UX behavior.

Medtronic · Luna

Rapid Research

A research cadence I created at PTC and scaled at Medtronic: a standing panel of 4 to 5 users across dozens of sessions, each scoped to testable questions.

Medtronic · Luna

Surgical Design Playbook

The sensory design language for Luna across screen, light, sound, and haptics.

Medtronic

Robotics heuristic framework

A custom framework for evaluating Human/Machine interfaces against human and robot interaction heuristics.

Adidas Group · Reebok

Material and creative direction

Led product innovation, CMF (color, material, and finish), and the seasonal design direction the whole category designed against. I also brought data into footwear design, a process adopted across the organization.

Range

Craft forged in automotive. Leadership learned in soft goods.

Two things that still shape how I lead: an eye trained where every surface gets scrutinized, and a category I built from zero while running the org that scaled it.

Honda R&D · Acura

Craft origin

Came up as an automotive designer at Honda, the toughest school for craft. Set the direction for Acura's surface language, Keen Edge, on the NSX concept, and designed the production MDX under full safety regulation.

Patents

2 design patents

Granted at Honda for vehicle design, part of a portfolio that now spans 8 granted and 6 pending across automotive and medical robotics.

Reebok · adidas Group

A $75M category from zero

Created the Nano, a 0 to 1 idea for a sport nobody was designing for. It grew into a franchise worth about $75M a year, around a third of Reebok's global footwear.

Global Design Head · a decade

The org that scaled it

Led a multidisciplinary team of about eighteen, with six to twelve direct reports, and owned the full people cycle: hiring, growth, critique, leveling, and advocacy in calibration.

Patrick Lukasak

About

A design leader who still designs.

I am a hands on design leader with twenty years across physical and digital products. I came up as an automotive designer at Honda, where every line, surface, and material got scrutinized. That is where my craft was forged. I became a leader in soft goods at Reebok and adidas, and that is where I learned the thing that still drives how I run a team: a design group only matters when its work ties directly to the commercial success of the business.

Today I lead UX for Medtronic's surgical robotics platforms. I grew that design team from two people to nine, and I still design. That is the part I protect. I am a player coach, not a manager who left the craft behind. I set direction, grow designers, and tie design to the business, and I am still in the work when it counts.

What I do best is take an ambiguous, first to market problem and turn it into something people can trust and use, whether that is a shoe, a factory floor dashboard, or a robot a surgeon relies on inside a patient. I hold 14 patents, 8 granted and 6 pending, across Honda and Medtronic, and I trained at Art Center College of Design and Rutgers.

AI is part of how I work. I used it to build OSDS, built two agents to help with Luna research, and design Luna with AI as a core feature. Outside my own team, I mentor in Medtronic's WISE program for women in science and engineering.

How I lead

Lead from inside the work

At PTC I built the operator dashboard and action tracker myself while directing four designers on the rest.

Win small, then scale

On Hugo I started with the 30 most triggered alerts and proved it at zero risk. Once the team saw the result, they bought into the full rebuild.

The human stays the expert

On Luna the system suggests and the clinician decides. Trust is the product.

Simplest version first

On DPM the richer charts tested poorly. Now I test the simplest version first and add complexity only when it earns its place.

Design has to move the business

At Reebok I built team structure and headcount around commercial priorities. The Nano became a $75M category.

Outside the team

Medtronic · Closing keynote

Broadcast to 90,000 employees

Closed an organization wide Medtronic conference with the Luna design story: how and why we use AI, AR and VR in the software.

Medtronic · WISE

Mentor for women in science and engineering

Mentor in Medtronic's WISE internship program, guiding early career women into science and engineering.

Training

Certifications

NN/g User Research. Design Thinking, Stanford d.school. Certified SAFe Agilist. Design for Manufacturing, Black Belt.

Writing · LinkedIn · November 2026

Designing AI overlays people can trust in medical devices

A practitioner's perspective on attention, uncertainty, interruption and human control.

Read the article →

Writing · LinkedIn · January 2026

Newton's connection to great UX design: There is Beauty in the Proof

User experience design and the scientific method.

Read the article →