Operational data only creates value when people can act on it.
Fruit assessments combine size, colour, defects and scan status inside time-sensitive packhouse workflows.
I lead the UX team at Hectre, bringing business understanding and deep human insight to Fruit AI — with hands-on work that includes a design system built entirely with Claude.
Fruit AI sits inside complex operational workflows for growers and packhouses.
Leading the UX team and building a design system entirely with Claude.
Original mobile and desktop interfaces, plus a documented design system built entirely with Claude.
Fruit assessments combine size, colour, defects and scan status inside time-sensitive packhouse workflows.
The experience must support interpretation without presenting incomplete AI output as a complete answer.
Processing, failure, unavailable data and completed assessments are visibly distinct across the interface.
Claude accelerated the system build. I defined its structure, product context, review criteria and quality bar.
A design system built entirely with Claude, with original interface work across mobile and desktop.
I built Hectre’s design system entirely with Claude: a documented foundation spanning tokens, atoms, components and modules for dashboard and mobile experiences. The system includes composition rules, interface states, accessibility decisions and resources for AI-assisted design.
Explore the design system Explore the uploaded offline design system directly in this portfolio.I bring the business context, human understanding and design judgement. Claude supports the creation process.
My approach to AI design ↗The design work turns AI-assisted creation into a governed product foundation: named tokens, dependency rules, reusable patterns, documented states and explicit review guidance. Speed is useful; coherence and human accountability make it scalable.
Size distributions, colour grades and defects sit together in one assessment view. The interface also distinguishes processing, failed scans, missing scans and unavailable data, so an incomplete result is not presented as a complete one.
A daily overview brings fruit received, recent loads, size and colour views, search and filtering into a mobile layout.
The sign-in concept applies the petrol and orange visual language to account access, with visible recovery, account setup and language options.
Shared Claude projects may require access. Original screenshots are shown above.
My starting point is the business problem and the people doing the work. From there, I ask where AI can help them understand, decide or act more effectively.
Start with the people, decisions and constraints behind an operational workflow.
Connect complex signals with an experience people can interpret in context.
Keep human responsibility, trust and control visible in the experience.
Hectre is a Fruit AI company helping growers and packers scan, document and manage their yields. Its operational context includes reliable fruit data, storage management, packouts and grower returns. This makes the relationship between information and real-world decisions central to the product experience.
As Product Design Manager, I lead the UX team. I bring experience in research, product vision, design systems and team leadership to an organisation with a strong AI focus. My perspective connects the detail of an interaction with the wider product and business context.
The load-assessment experience brings distributions, grades and defects into one view while making processing, failure and missing-data states explicit. The design decision is deliberate: a user should understand the quality and completeness of the available information before acting on it.
I use AI across research synthesis, prototyping, systems and workflows, but I do not outsource product judgement. At Hectre, Claude helped turn defined foundations into a comprehensive design-system artefact; the business context, architecture, accessibility review and decisions about what belongs in the product remained human-led.
Explore my AI practice ↗