Thoughts, links, videos, and things worth saving.
May 2026
Built a product from concept to launch in four days with AI assistance. That used to be a four-month project. Something has genuinely shifted.
Went from the Claude Code free tier to a 20x Max account in three days. No regrets.
April 2026
I've been using Gary Tan's gstack skills stack with Claude Code. It's one of those workflow things that feels like overkill until suddenly it isn't.
Three months ago I was writing everything by hand. Now I talk, Plaud captures it, AI shapes it into something usable. The bottleneck moved.
The Plaud Note is one of those devices that sounds gimmicky until you actually use it. I capture more now than I ever did trying to type notes in the moment.
Started building an internal system for True Build. The off-the-shelf tools were fine. A system built around how we actually work is better.
I'm less interested in AI as a novelty and more interested in what it lets a small team do that used to require a much bigger one.
n8n has quietly become one of the more useful things in my toolkit. Automations I used to build custom code for now take an afternoon.
March 2026
Zoho Books and Bigin aren't glamorous but they work. Most of the tooling decisions that matter are boring.
The instinct to start building before the thinking is clear is the most expensive instinct in any project. Software or physical.
There's something clarifying about working across construction sites and codebases in the same week. Both reward the same thing: clear thinking before you touch anything.
The best projects aren't the ones with zero problems. They're the ones where problems surface early and get dealt with properly.
Design is a problem-solving discipline. The aesthetics are the output. The thinking is the work.
Good delivery means protecting the design intent through every practical decision, not just the first one.
February 2026
Every renovation has hundreds of hidden decisions. A socket location, a tile joint, a door height. Each one feels small. Together they determine whether the project feels considered or chaotic.
There's a big difference between looking professional and operating professionally. A lot of businesses have learned the first one. Far fewer have built the second.
Documentation isn't admin. It's leadership. The record is what keeps a complex project calm.
"Flexibility" gets treated as a virtue on projects. It's often just a lack of structure wearing a nicer name.
Clients often don't need more options. They need fewer, better-considered ones. Too many options is the designer outsourcing their judgment to the client.
Most renovations don't go sideways because of bad design. They go sideways because decisions were vague and nobody was keeping the record.