Blog
Notes on engineering leadership, AI adoption, and running technology organizations that outlast the person who built them.
1 post · latest: What Claude's New Memory Controls Mean for Your Firm's Confidentiality Obligations
Everything we publish, in one place — writing on the work, a weekly read on what changed in AI, and the reference sheets we got tired of rebuilding from scratch every time a client asked.
Notes on engineering leadership, AI adoption, and running technology organizations that outlast the person who built them.
1 post · latest: What Claude's New Memory Controls Mean for Your Firm's Confidentiality Obligations
A once-a-week read on what actually changed in AI and what it means for a professional-services firm — platform news, an industry spotlight for legal and finance, and one thing worth trying. No jargon, no hype.
6 issues · latest: No. 6, September 15, 2026
A quick-reference guide to the two Anthropic admin consoles — claude.ai organization settings and the platform.claude.com Console — covering roles, SSO, spend limits, usage reporting, rate limits, and the canonical doc link for each.
Which Claude model to use and why — how the Haiku, Sonnet, Opus, and Fable tiers are structured, how to choose between them, where model selection actually happens in each product, and the canonical Anthropic doc links.
One email a week on what changed in AI and what it means for your firm. You can read the back issues first — the signup form is at the top of the archive.