Hi, I’m Katherine 👋🏻

Hi, I’m Katherine.

I’m a principal-level product design lead / Design Architect who designs AI product systems: the behaviours, workflows, and foundations that make intelligent experiences reliable, explainable, and shippable.

I work at the intersection of AI system behaviour, interaction design, and design systems. That means I’m not just designing screens, I’m defining how the system plans, retrieves, uses tools, handles uncertainty, and fails gracefully. If I don’t understand the behaviour, I’m not designing I’m decorating.

I’ve spent 10+ years designing complex digital platforms across enterprise and high-constraint domains (tech, engineering, finance, mobility, crypto). I’ve worked in both startups and top enterprises, so I’m comfortable moving fast while building foundations that scale.

Alongside delivery, I completed Human–Computer Interaction (HCI) for AI Systems Design at Cambridge, where I focused on how AI reshapes human decision-making, trust, and interaction patterns, especially in high-stakes contexts.

I’m based in Bristol, UK with my family and two cats. Outside work: books, exploration, and anything that teaches me a new mental model.

How I work:

  • Define system behaviour before UI (states, rules, failure modes, quality bars)

  • Prototype early with real constraints (prompts, data, tool flows)

  • Build scalable patterns (design systems + governance)

  • Translate across teams (engineering truth + stakeholder clarity)

Trusted by

My recent projects

Solving complex problems, creating meaningful user experiences, and driving real business value.

AI Product Design Lead — Avalere Health (SenseAI Ecosystem)

I lead system-level design for SenseFamily, a suite of AI products powering Avalere’s knowledge and decision workflows. My core focus is not UI polish — it’s designing how the AI system behaves: how it retrieves evidence, uses tools, handles uncertainty, and earns trust in high-stakes, compliance-driven environments.

I operate as a Design Architect / Creative Technologist across the full stack of product thinking: interaction models, orchestration flows, quality expectations, and scalable patterns that multiple teams can ship consistently.

2025-Now

  • AI behaviour & conversation design: multi-turn patterns, system states, clarification loops, “can’t answer” handling, and safe escalation paths

  • Evidence-first UX: citation behaviour, source inspection, traceability, and “why this answer” explanations designed for trust and defensibility

  • Orchestrated workflows: intake → plan → retrieve → draft → verify → refine → publish, with progress visibility and step-level error recovery

  • Quality bars & evaluation signals: what “good” looks like (consistency, citation coverage, reliability, latency tolerance, safety behaviours) and how teams measure it

  • Governance-by-design: auditability, policy constraints, and human-in-the-loop checkpoints embedded into the experience (not bolted on)

  • Ecosystem consistency: shared patterns and reusable interaction components across multiple Sense products, preventing “one-off AI UX” drift

  • Cross-discipline alignment: deep partnering with Engineering, Data Science, Clinical/SMEs and Compliance — translating constraints into buildable product behaviour and translating technical reality into stakeholder clarity

Design Lead @ BT — Global Fabric (Network-as-a-Service platform)

At BT, I led system-level product design for Global Fabric, a programmable connectivity platform that combines ordering, configuration, provisioning, observability and lifecycle management into a coherent enterprise experience.

My work went beyond UI. I shaped the interaction model, information architecture, and system behaviours that made the platform understandable and scalable for both users and engineering teams.

2022-2025

  • Service design & end-to-end flows: mapped the full service lifecycle (discover → quote/order → provision → operate → change → troubleshoot → retire), aligning user journeys with backend provisioning realities and operational constraints

  • Information architecture & mental model: defined the object model (e.g., sites/locations, ports, circuits, policies, cloud connections) and structured navigation, hierarchy, and cross-linking so users could predict where things live and how they relate

  • System behaviour & state logic: designed interaction rules for complex states (pending/provisioning/active/degraded/failed), error recovery, confirmations, and safe “undo/rollback” moments — so the UX matched how the network actually works

  • Cross-squad alignment: acted as the bridge between design, PM, engineering and architecture — translating technical constraints into product language, and stakeholder intent into buildable system requirements

  • Design system & governance: established a scalable component and token foundation (variables, responsive rules, reusable patterns) so multiple squads could ship consistently without design drift

  • Platform standards: codified patterns for dashboards, tables, filters, status indicators, and configuration flows — making future capabilities composable rather than re-designed every time

Senior Product Designer — Société Générale / Automotive & fintech platforms

I designed and improved complex B2B/B2C platforms across mobility and financial services — focusing on usability, scale, and real operational constraints.

I led research and iterative delivery, turning legacy complexity into streamlined workflows, clearer information design, and measurable improvements in user efficiency.

Me in Real Life

I’m curious, adventurous, and always learning. I like hard problems, strong opinions backed by evidence, and building things that actually work.

Outside of work you’ll find me reading, exploring new places, or riding horses, usually with a new tool or side project in progress.

Turning AI ambition into shippable, trustworthy systems — with design foundations that scale.

Turning AI ambition into shippable, trustworthy systems — with design foundations that scale.

Turning AI ambition into shippable, trustworthy systems — with design foundations that scale.