Imprints

The stake

A shopping question is rarely one question. People arrive with a vague goal, a budget and a fear of choosing wrong. A chat box alone does not solve that.

The call I made

I treated Rufus as a service, not a chat window. The design work covers the whole journey, from first use and onboarding through discovery, checkout and order changes, so help appears at the moment of doubt.

What I lead

• Service design strategy for Rufus, shaping multimodal discovery and personalized recommendations.
• Global programs that align AI science, product, engineering and research on one roadmap.
• Checkout, onboarding and cancellation experiences that remove friction from first use to order changes.
• Service design frameworks that combine behavioral research, machine learning insight and journey mapping.

The stake

Shopping by voice has no shelf to browse. The assistant has to know enough about you to be useful, and ask little enough to stay out of the way.

The call I made

I made the case that Alexa shopping should feel personal, not generic, and aligned product, science and engineering around personalized experiences instead of one size fits all features. I wrote the strategy document that set that direction.

What I led

• UX strategy for Alexa Shopping, advancing conversational and multimodal experiences powered by natural language understanding and predictive personalization.
• Service design models and journey frameworks that connect voice, visual and mobile shopping into one experience.
• Design principles, governance and success metrics for AI driven products.
• Strategy sessions that translated AI capability into clear narratives and aligned teams on vision, experimentation and roadmap.

The stake

On a busy ward, a critical chest X ray can wait hours for a radiologist to read it. AI could flag the urgent ones first, but only if the radiologist trusts what it says.

The call I made

I stopped treating trust as one problem. Radiologists meet AI in three different mindsets, so the design had to respect all three: show the AI's confidence at all times, let the clinician accept or reject every finding, and always allow an off switch.

What I led

• Design research on how clinicians form trust in AI, and the interaction model that came from it.
• The experience of showing newly developed AI inside existing radiology workflows.
• A unified design system across the Edison platform, recognized for craft by the awards below.
• A research, design and design operations organization of 40.

The stake

Drivers buy map updates through each automaker's store. Every brand, language and region had its own habits, and the experience had to hold together while the business grew.

The call I made

I built the operating system before the screens: one information architecture for the whole store, a testing culture that let every team learn from live customers, and an outsourcing model that added capacity without losing quality.

What I led

• Information architecture. One structure for the full HERE Store, used by development to code, QA to write test cases, analytics to track, and marketing to see product vision.
• A culture of optimization. A six step testing framework and a first phase of 12 mini tests, with every team on board.
• Outsourcing, built from scratch. I interviewed vendors, shortlisted a few, set up agile, sprint demos and QA from day one, and grew it into a whole outsourced department.
• UX vision and strategy. Presented to senior management, development, QA, operations and marketing, with an action plan for the next six months.

The stake

Flexera sold many point products one at a time. The people using them wanted one complete view of a complex IT environment, so they could manage technology spend and risk before problems arrived.

The call I made

I defined the UX vision for the whole company and drove it to completion. Instead of redesigning products one by one, I framed the work as a maturity journey customers move through, and gave every team five principles to design by.

What I led

• A UX audit of the Flexera One platform and an assessment of competitors, turned into a prioritized roadmap.
• Four personas, from the executive who buys to the operator who lives in the product.
• A long list of use cases narrowed to a priority set, then tested through customer and partner interviews.
• Related work: a SaaS management retention study, a help and support redesign delivered in two weeks, and a design system any team could ship with, without guidance.