Articles
Operator notes.
Problem-first writing on AI, automation, and the process underneath. Each piece names a real bottleneck, exposes the broken layer nobody owns, and shows what the working system actually looks like.
A legal AI agent reviews contracts in minutes. It can't fix a clause taxonomy that drifts.
Legal AI agents now extract and review clauses end-to-end. But the agent inherits your clause taxonomy — and when one clause is named three ways, it misses it, silently.
7 min read → HealthcareAI medical coding is only as accurate as the note underneath it
AI medical coding now auto-generates ICD-10 and CPT codes from the clinical note. But the code inherits the documentation — and a confident code on a thin note books a denial faster.
7 min read → Financial servicesAI lending agents can run the whole loan. The credit decision still rides on the loan file.
AI lending agents now automate the full loan lifecycle — but the credit decision inherits an un-owned loan-file data contract. Fix the file, then run the agent.
7 min read → Enterprise AIAI cost controls just shipped. A dial is not a budget owner.
Anthropic added AI cost controls to Claude Enterprise in July 2026 — caps, alerts, model entitlements. Useful. But a control enforces a budget; it doesn't own one.
7 min read → Customer serviceWhy AI customer service agents are only as good as the record behind the answer
AI customer service agents lifted CSAT to the #1 improved metric in 2026. The catch: the agent makes a public promise on internal entitlement, refund, and order data that's often wrong.
7 min read → Enterprise AIAgentic arbitrage: why Gartner's $234B SaaS warning is really a legibility test
Gartner says agentic arbitrage puts $234B of SaaS spend at risk by 2030. The real signal: which of your processes an agent can run without the screen.
8 min read →