OnEarly is a healthcare startup led by Andrew Barnes (CEO and co-founder of Go1) with the mission to revolutionise how Australians manage their health. The platform combines traditional healthcare records, blood tests, MRI scans, clinical data, with AI capabilities that surface patterns, track changes, and provide personalised insights. I joined as the founding designer after leaving CarExpert, shaping the product vision during its earliest stage while also establishing Insitu Digital.
OnEarly was created to close a critical gap in healthcare: the lack of a unified, patient-centric platform that connects data, insights, and action. Rather than siloed test results, OnEarly provides an all-in-one product — a hub for medical records, peer-reviewed articles, and an AI health companion.
As the founding designer, I worked directly with Andrew and the engineering team to:
The goal was ambitious: Make health data not only accessible but actionable, while ensuring trust and clarity in a space often weighed down by complexity.
This was my first project with AI as the core product driver, and it quickly became clear that its role in healthcare went far beyond novelty. OnEarly was built around the idea that AI could act as both a companion and a lens — helping patients understand their own health data, while supporting practitioners with context and insights.
Some of the possibilities we explored included:
OnEarly launched its MVP in partnership with OneMRI, an Australian service offering full-body MRI scans. It currently powers the backend for OneMRI and will soon expand to other providers. The next iteration will introduce a consumer-facing experience, allowing customers to set up and purchase tests and scans directly in OnEarly without jumping between providers.
From a technical perspective, this was a massive challenge and talking point. The platform needed to be broad enough to support any type of practitioner/admin, while remaining clear and simple for users. Building OnEarly with Expo meant designing for both web and app concurrently, ensuring a consistent, scalable experience across devices from day one.
Because we were building OnEarly from the ground up, I had the rare opportunity to set the design system foundations from day one. My goal was to ensure the system was scalable, flexible, and developer-friendly. I worked closely with the lead developer to make sure components mapped cleanly back into Storybook, supporting both light and dark modes from the start. This was the first time I’d built a DS without have anything to go off (it was challenging and fun). It pushed me to think not just about design, but also about how the system would live, evolve, and stay maintainable for the long term.
In parallel, I worked in Miro to wireframe the product, collaborating closely with Andrew to unpack exactly what was required. We mapped out both consumer and practitioner flows, drilling down into the smallest interactions. Structurally, we leaned on a model similar to Facebook Groups:
This concurrent approach meant the system and flows informed each other. The design system gave us consistency and guardrails, while the wireframes stress-tested whether the components we were building were flexible enough to support real-world use cases.
This meant I could move forward into high-fidelity design with confidence, knowing the foundations were solid and scalable.
Designing at the frontier is challenging but worthwhile: Being the founding designer on a project where AI and healthcare intersect sharpened my ability to design for ambiguity and scale.
Vision vs. execution: Conceptual projects demand bold vision, but also pragmatic execution to prove value early.
AI as a partner, not just a gimmick: The real opportunity lies in making AI genuinely useful, surfacing insights, not just chat.
Startups are momentum games: OnEarly reinforced how fast a small, backed team can move when vision, funding, and timing align.
