Thoughts on AI, engineering, and building
Practical writing on AI agents, automation, and the engineering behind them, for the people who run the business.
Harnesses & Toolingdeep-dive
How the Model Context Protocol's 2025-11-25 revision handles authorization and long-running work, and what that means for CRM and ERP integrations.
· 4 min read
Economics & Pricingexplainer
Three levers that cut LLM spend without cutting quality: route by task, cache repeated context, and batch what is not urgent. Real list prices included.
· 4 min read
Enterprise Adoptionbrief
Lock-in sits in several layers. A checklist of what to abstract (API layer, prompts, evals, embeddings, data) and what vendor docs say about switching.
· 2 min read
Hardware & Infrastructurehardware-breakdown
What Nvidia announced at Computex 2026, which claims are vendor estimates, and what the specs mean for teams planning inference capacity.
· 4 min read
Enterprise Adoptionchecklist
A week-by-week checklist for the first month of an AI pilot: scope, baseline, simplest build, real-data testing, and a clear go or no-go.
· 2 min read
Agentsexplainer
An agent has three kinds of memory: the context window, saved session state, and long-term memory. Here is how they differ, what they cost and what to govern.
· 3 min read
A deep dive into architecture patterns for production AI agent systems.
Enterprise Adoptiondeep-dive
Customer support is one of the better-studied uses of generative AI. What one large field study and a vendor's own numbers show, and how to design for retail.
· 4 min read
Economics & Pricingnews-analysis
Alphabet, Microsoft, Amazon and Meta reported Q1 2026 results. What they spent, what they guided, and why component pricing is now part of the story.
· 4 min read
Harnesses & Toolingexplainer
A production prompt is a versioned specification with real examples, a schema and an evaluation set. What the vendor docs say to build before you tune wording.
· 3 min read
Harnesses & Toolingexplainer
An AI agent that fails rarely says why. Tracing records each model call, tool call and handoff so you can find out. Here is what to capture and what to watch.
· 3 min read
Economics & Pricingchecklist
A practical framework for deciding whether to build, buy or partner on an AI capability, with the evidence and its limits.
· 3 min read