The research workflow that replaced my analyst.
Half a day of client prep, cut to forty minutes — and the forty minutes are better. The exact Claude + Obsidian pipeline, including the two steps I had to throw away.
AI workflows documented from inside a working consulting practice — not tool reviews from someone who opened a free trial yesterday. What survives past the novelty. What I cut. What a senior professional should actually copy.
The exact pipelines I run, step by step — prompts, tools, and where I still intervene by hand.
Choosing between tools that look identical on a landing page — decided by living with them for a year.
Longer essays on using AI as a thinking instrument, drawn from supply-chain systems work.
The machine is fast at "here are ten things." You are paid for "here is the one that matters."— from File No.01 · "The research workflow that replaced my analyst"
Finally, someone writing about AI tools who has actually had to put output in front of a client. The failure annotations alone are worth the price of the template.
I've subscribed to a lot of AI newsletters. This is the only one I look forward to — because nothing in it has ever felt written to fill a quota.
The line — automate gathering, never judgement — reframed how I use Claude for client deliverables. Three months later it's still the rule I quote in my own head.
Half of what gets called 'thought leadership' on LinkedIn is people summarising a free trial. This site is what the real thing looks like.
No schedule. No filler. No sponsored placements. A new build goes out only when it's worth your ten minutes — usually every 2–4 weeks.
"I'll only send you something I would have written even if nobody paid me to. The day that stops, the newsletter stops."