What I'm actually doing.

In build · two products

GreenR

The one I've wanted to build for years. A rated, searchable directory of recruiters — filter by city, industry, salary band and work style, then match with the ones who'll actually advocate for you. It exists because I applied for hundreds of roles, got screened out by ATS filters and AI, and it took one human recruiter to change the outcome. Lovable and Supabase under the hood. Candidates stay anonymous to recruiters until they explicitly approve a reveal — a deliberate design decision, not an afterthought.

Overran its deadline

DropIn

A travel gym drop-in app. I gave it a six-week public deadline of 7 June and missed it. Search works, gym pages render with real drop-in pricing, and the core flow holds — but it isn't live, and I'm not setting a new date until I can hold one. The build log has the honest version.

The unglamorous half

Infrastructure, not features

Where most of the hours actually go: transactional email through Resend so confirmation emails send at all, DNS records, anonymising personal data, chasing a registry hold on a .com.au domain, and hunting dead links. Vibe coding gets you a working screen in an afternoon. It does not get you a product.

New series

Dropping In

A content series running across LinkedIn, YouTube, and Instagram (@2mtechtalks). The thesis: the unsexy middle of building — governance, infrastructure, the boring choices that don't go viral — is where the actual leverage lives. Currently proving it the hard way, in public, with two half-built products.

Ongoing

2-Minute Tech Talks

Still going. Slowly closing the gap between corporate training videos and something a busy operator would actually watch.

In progress

AI fluency for operators

Going deep on Anthropic and Claude certifications. The interesting question isn't "can AI do this?" — it's "what does an operator need to understand to deploy this safely?" Different question, different answer.

Sat with it

Impostor phenomenon

An Atlassian Unstoppable You session that stuck — thank you Zuli Posada, Brooke Grant, Lara Scharenguivel and Marion D'Cruz. The line I keep coming back to: share your failures, not just your wins. Coming from a non-technical background, the temptation is to stay in the part that feels like magic and skip the part that feels like admin. Showing up for the boring part is still showing up.

Learned the hard way

Shipping is a muscle

Building something low-stakes and genuinely fun first is what got me out of my own way on the project I'd been sitting on for years. Perfectionism doesn't produce a launch. It produces a very well-considered nothing.

CrossFit

Productive DOMS only

The DOMS-as-metaphor for production failures keeps writing itself. Soreness from a smart program is the cost of progress. Soreness from a bad deploy is the cost of skipping the program.