Tristan Clarkson Contact

Design Operations LeadDarwin, part of MarshNorth West England, UK

TristanClarkson

Open to Lead UX and product design rolesSay hello

I make complicated systems understandable to people and buildable by teams.

Regulated services. Enterprise platforms. AI that has to admit what it does not know.

I'm Tristan. Around fourteen years in UX, UI and product design, and for the last few I've led design operations at Darwin, part of Marsh.

I design the product, prototype the hard parts in code, and build the standards that keep quality up. I'm looking for a Lead UX or product design role where the problem is difficult.

Download CVRead it online

The reel

About twenty-five seconds of real artefacts. No sound.

Skip the reel

Showreel

Three rooms.One habit.

Find what is actually hard. Then make it workable.

Representative Darwin decision model: a need passes five questions about standard UX, added value, confidence, autonomy and uncertainty, ending in a standard pattern, assistive AI, AI with human verification or not appropriate for AI.

Darwin, part of MarshShould this be AI at all?

Darwin assistant response that says it cannot confirm the answer and explains what the evidence covers.
Darwin human handoff screen explaining why specialist support is appropriate and which context can be shared.

Design explorationsUncertainty and handover, designed as states.

Luna dashboard showing documents, profile, mileage, banking, tax and VAT services.
Luna home on a phone, with the task that needs attention first.
Luna transactions on a phone, with totals and a classification action.
Luna VAT screen on a phone, showing return status and figures.

Luna, IN-SYNCPosition first. Then the next action.

Rejoin active journey with platform and wayfinding information.
Rejoin explaining a cancelled journey and preparing recovery.
Rejoin recommending one checked recovery route.

Rejoin, a designed conceptOne checked next action. Or an honest no.

Now thelong versions.

DarwinLunaRejoin

Intro

Three rooms.

A practice across an organisation, a product shipped end to end, and decision design under pressure. Open one and it unfolds.

Darwin, part of Marsh · Design Operations Lead · 2022 to present

Whether AI belonged at all. Case study: making AI repeatable across Darwin, part of Marsh.

AI was arriving in chat, summaries and search, and every team risked inventing it again. I connected strategy, language, governance and components so responsible AI could become repeatable.

Evidence: a framework, patterns and design explorations. No adoption figures are claimed.

Darwin benefits page with a contextual AI policy summary, a verified source and follow-up actions. Darwin AI framework: a coherent system for responsible, consistent and scalable AI. Darwin assistant response that states it cannot confirm the answer and explains what the evidence covers.

IN-SYNC Group · Lead UI and UX Designer, web developer · 2019 to 2021

Position first. Then the next action. Case study: Luna, digitising a tax service without handing the tax to the customer.

Making Tax Digital changed a service customers had relied on IN-SYNC to perform. Luna showed where things stood and what only the customer could do, and kept the tax mechanics with the service.

Evidence: Luna launched. No measured results are available, and the case says what I would measure.

Luna dashboard showing documents, profile, mileage, banking, tax and VAT services.
Responsive Luna home showing the task that needs attention and the service areas. Responsive Luna transactions overview with classification actions and totals. Responsive Luna VAT screen showing return status and calculated figures.

The screen matters. So does everything that makes the next screen better.

  1. Find the real constraint

    A difficult interface often points to a service, model or decision problem upstream.

  2. Prototype the hard parts

    Working behaviour answers questions that static frames flatter or hide.

  3. Make the decision visible

    Research and reasoning matter when a reviewer can see what changed because of them.

  4. Leave a system behind

    A pattern, standard or shared route can keep improving work after its author leaves the room.

More about how I think

Say hello

Have a difficult product, or a practice that needs firmer footing?

Tell me what is hard about it. I'm open to conversations about Lead UX and product design roles, design operations, AI experiences and regulated enterprise products.

jeremeze@gmail.com