16 years of design experience, the last eight on software products at an early-stage start-up, a fast-growing scale-up and an ASX-listed company. I mentor and hire designers, set direction for the products I work on, and use AI coding tools to work closely with engineers, from code prototypes to polished code changes.
I believe the most important design decisions are about framing the problem, so I help teams agree on what they are solving before they start designing.
Work samples
Examples below are from my time working on Ignite, REA's dedicated platform for real estate agents.
Dec 2025–Mar 2026
A new home for Ignite
A senior leader handed us a simple, open brief: make the Ignite home page better, on a tight deadline before Christmas. I brought two senior designers and a researcher together, got us sketching early, and pushed until we landed on three distinct directions.
Testing with customers uncovered the real problem: people couldn't easily reach their relationship manager, and listing performance wasn't visible enough.
Aug 2025–Present
AI-powered campaign management
Built by a 'tiger team' with a fixed budget and one open question: how could Ignite make use of LLMs? It came together just as tools like Claude Code got everyone talking about AI-assisted coding, and that momentum pushed the question up the priority list.
The result was Campaign Assist, which helps agents prep for vendor catch-ups on the go: surfacing comparable properties, campaign performance and recommended next steps like 'listing bumps', all inside a chat interface.
Alongside the product work, I helped a number of designers across our business unit get up to speed with AI-assisted coding, building prototype playgrounds with our design system baked in.
2022-2025
Turning REA's property data into a new revenue line
Ignite's value to agents had always been reporting on vendor-paid advertising (listings on realestate.com.au). The opportunity was to find another source of revenue, one that didn't depend on residential advertising. Built for agents.
A subscription only works if there's something worth paying for and every agent has to price a property before they list it (known in the industry as a CMA or Competitive Market Analysis). REA already held the data to do this well. But this was a market with two established players: CoreLogic and Domain's Pricefinder. Agents don't swap out tools they trust on a whim, so whatever we built needed to earn that switch.
Aug 2026
Meeting (Real Estate) Agent's where they're at
A month-long hack exploring whether Campaign Assist's chatbot could work outside Ignite entirely. We proved it could run inside WhatsApp, then extended the idea to leads, letting agents get notified and respond to enquiries from a channel they already live in.
We shipped a working coded prototype, tested it with real customers, and demoed it at REA's Industry Leaders Forum, a chance to show where AI could take the product next.