Building in public
AI automation, apps, and marketplaces — rough draft to shipped. Watch the process, take the playbook, or hire the builder.
Pipelines and tools that do the repetitive thing reliably — not just in the demo. The boring layer where the real value lives.
Full products with real users and real money moving through them. Payments, splits, the parts that quietly break if you get them wrong.
A 3D printer and a laser running most days. Because not every problem lives on a screen, and the method is the same either way.
A marketplace where racing photographers sell shots to teams. Every sale splits automatically between photographer and platform through Stripe, with print-on-demand as an option. The hard part was never the photos — it was getting the money to move correctly, every single time.
Visit Raceframe →A music player I built for my uncle, who has schizophrenia and lives in a care home. Plug in a USB drive, it copies the songs, shows a list, hit play. That is the whole app — built so it cannot be broken. The best engineering I do usually looks like removing things until what is left can’t fail.
A system that generates a full short-form video on command — keyframes, motion, native audio, final assembly — with consistency contracts and a quality check at every step. Built as reusable skills, so the next one takes minutes instead of hours. Most AI content fails because people just generate. The win is the scaffolding around it.
Got a repetitive process you keep meaning to automate and never get to? That is the stuff I build — content pipelines, internal tools, small systems that quietly give you hours back.
Not a pitch deck. Tell me the annoying repetitive thing and I will tell you if it is buildable.
hello@shoplog.buildEverything I make, with the wiring shown — what worked, what I would redo, what it cost. No spam, no fluff.
↑ placeholder — paste your Beehiiv embed here