Notes from
the work.

What I learn building automation and software for businesses. 233 posts, free to read, no sign-up.

The AI Implementation Opportunity Report 2026

12,000 words and 80+ verified sources: pricing benchmarks, opportunities by sector, GEO trends and the root causes of small-business AI failures that most reports miss.

All 233 AI agents and tools 31 AI strategy 75 Automation 68 Software and development 37 UK business and governance 22

Showing 1 to 7 of 7 posts

Software and development22 Feb 20264 min read
MCP Is the Most Important AI Standard You Have Never Heard Of Model Context Protocol is the USB-C of AI. 97M monthly SDK downloads, adopted by OpenAI, Google, and Microsoft. Here is what business leaders need to know.
Software and development22 Feb 20265 min read
The Number One Reason AI Projects Fail Is Not the AI, It Is Your Legacy Systems 46% cite legacy integration as the primary challenge. 60% of AI leaders say it is the top barrier. Here is how to fix it without replacing everything.
Software and development20 Feb 20269 min read
The Model-Agnostic Future: Why the Best AI Tools Let You Choose Your Own Model Vendor lock-in is the hidden cost of proprietary AI tools. The smartest teams are building model-agnostic, and here's why you should too.
Software and development20 Feb 202610 min read
The Integration Tax: Why Connected Systems Beat Individual Tools Data silos and manual data transfer between tools cost more than most teams realise. The integration tax is the hidden drain on your operations.
Software and development17 Feb 20268 min read
Small Models, Big Results: Why Domain-Specific AI Is Beating GPT for Business Tasks Fine-tuned small models outperform general-purpose LLMs for specific business workflows while costing 10-50x less.
Software and development17 Feb 202610 min read
Small Language Models Are the Enterprise AI Story of 2026 AT&T's CIO called it right, fine-tuned small language models are the real 2026 story. Edge AI, privacy, and 10-50x cost savings are making SLMs the smart play.
Software and development17 Feb 202611 min read
Why One AI Model Isn’t Enough Anymore: The Multi-Model Strategy Perplexity's Model Council runs Claude, GPT, and Gemini in parallel to cross-validate answers. This is the future. Here's how to implement it.

Want this done for your business?

One paragraph is enough. I reply personally.

Bring me your worst workflow