Latest insights, tutorials, and thoughts on business, technology, and innovation
Field notes on building real things with LLMs. Occasional, practical, no hype.
A family-history app for an 81-year-old user, looked after by a headless Claude Code agent that answers his comments and ships fixes every hour. The rules that make that safe, and the bug it found on his first morning.
We built a local web app that sends trail-camera frames to Claude, bins the empties, and reports what the camera saw. Then we open-sourced it, and the privacy pass found things a grep never would.
We zipped our Claude Code setup into a 37 KB starter kit a friend can install in half an hour. Sorting the files into portable, needs-editing and not-portable was the real work.
Our TV kept buffering while the mesh app insisted everything was fine. Six days of continuous measurement found one failing node, after four wrong theories. Here is the method, which transfers to any home network.
We generated a full set of branded video-call backgrounds with an AI image model, and learned that consistency comes from one locked master image and add-only edits, not from better prompts.
I pointed an AI coding agent at a plan and said 'build this project'. Out came a complete multiplayer browser game you can play right now. Here is what a single prompt can really do, and the one thing that makes it work.
We built a Claude Code skill that audits any site with Google Lighthouse and produces a fix plan with one hard rule: it must never change how the project renders. The pattern of treating your architecture as a machine-checkable invariant is the bit you can steal.
We built a tool that watches a web app the way a person would, then produces a narrated, on-brand promo video from nothing but a URL — and the whole thing was built, and debugged, by an AI coding agent that could see its own video output.
We extracted 1,451 of our own prompts from Claude Code's session logs and counted what we kept retyping. The repeats became three new skills and a pile of instructions the agent now reads automatically.
We took a one-off security-audit prompt, tightened it, and turned it into a global Claude Code skill you can run in any project with /security-audit. Copy the code — or the prompt to generate it yourself.
The best material for a developer blog is already in your terminal — but letting an AI write it up means guaranteeing it can't publish on its own. Here's the draft-only pattern we landed on.
An AutosBuzz 2026 talk on how AI and LLMs are reshaping automotive marketplaces — the data, the precedent, and the strategic choice every marketplace now faces. Watch the talk, read the slides, and try the live build.
After writing an entire book on AI-assisted development with Claude Code, here are the seven insights that changed how I think about building software — and why the future belongs to developers who learn to collaborate with AI.
A practical guide to making your website content accessible to AI systems like ChatGPT, Claude, Perplexity, and other LLM-powered tools.
A complete guide to integrating Simple Analytics with Next.js. Learn how to build privacy-compliant analytics dashboards, solve common integration challenges, and create custom insights that drive business decisions.
How we migrated from WordPress to Next.js in just 8 hours, built a custom analytics platform, and achieved 75% performance improvements. A complete guide for businesses considering the switch.
How we transformed a standard 404 page into an engaging experience using VO3 video generation and a creative technical approach.
Discover proven sales negotiation techniques from Neil Rackham’s research on skilled negotiators. Learn how flexibility, clarity, and strong arguments lead to successful deals.
I've been testing AI for some time now. Have tried most of the 'non technical' tools but am now sticking with a stack of Lovable, Supabase, Netlify and Devlo AI with Chat GTP for when things get...
Using JSON to create a VO3 Flash2.5 Gemini videos is by far the best way If you’ve been experimenting with Google’s Gemini VO3 video generation model, you’ll know that the quality of your results...
My thoughts on how to future-proof your brand for the LLM Era. Every marketing team I speak to is still investing predominantly in Google: * SEO strategies fighting for Page 1 or Result 1 * ...
AI is moving fast. By using more detailed prompts, encouraging thoughtfulness and contemplation in GTP 4o seems to create much better article quality.
Don't have time to read this? Listen instead: (Or, I read a book: my thoughts on Dale Carnegie's classic, "How to Win Friends and Influence People.
Don't have time to read this? Listen instead: The timeline of Cazoo. Cazoo, a British online car retailer, was founded in 2018 by Alex Chesterman with a vision to revolutionize the used car market.
I might as well hop on the Cazoo bandwagon. Reading the Cazoo saga made me reflect on my journey with Carsnip, with the challenge of establishing a digital marketplace in the car industry.
MarketCheck USA Inc has entered into a strategic partnership with Hurricane Works, the consultancy company founded by the ex-CEO of Carsnip.com The collaboration will bring MarketCheck’s data...
The surprising sites not using https - leaving you wide open!
how to disable lightspeed bitnami banner in AWS
how to disable lightspeed bitnami banner in AWS
Field notes on building real things with LLMs. Occasional, practical, no hype.
