How I use AI to build products.

I use AI to help me deliver software faster. That includes complete websites and larger platforms. But I still have to understand the business, choose the architecture, and take responsibility for what gets shipped.

Start with the business.

With Clarity, the work started in meetings with lenders. They brought me in as their CTO for the months they needed, and I worked through how their lending facilities would connect to Shopify and other ecommerce stores.

Those conversations shaped the platform. AI helped with engineering, but understanding the requirements and making the decisions was my job.

Use the speed. Keep the judgement.

I built the platform in Go in under a month, including lender integrations, ecommerce integrations, validation and testing. AI was part of how I moved quickly. Experience was how I decided whether the result was right.

The same applies to a small marketing website. An agreed brief and modern tools let me move quickly, but the site still has to make sense for the business and be something the owner can use.

Being a good engineer still matters.

My view is that AI makes your existing engineering habits more consequential. If you know what to look for, you can spot a wrong assumption or a design that will be painful to maintain. If you don’t, generating more code can make those problems harder to see.

A convincing demo isn’t the end of the job. The integrations need to work, the important behaviour needs validation, and the business needs a product it can keep using.

I own what I deliver.

Whether it’s firmware, infrastructure, AI integration or a client platform, I’m responsible for the decisions and the result. AI helps me get there faster. It doesn’t take that responsibility away.