Library
Cases
Case-shaped writing by domain, built for pattern recognition rather than client theater.
Cases are generic by domain and specific by method. The point is to make the operating pattern visible: what was claimed, what the workflow required, what the source evidence supported, and what changed.
Clinic operations
Scheduling, documentation, staffing, queues, compliance, exceptions, and the local habits that determine whether a system is usable.
No clinic-ops case published yet.
Data work
Source coverage, metric definitions, exports, reconciliation, and the unglamorous work that decides whether a dashboard can be trusted.
Sample: clinic data quality audit
A case-style post that uses charts to show where an operating report can and cannot be trusted.
AI buildouts
AI systems inside healthcare workflows: what they prepare, what they own, what needs review, and where the demo stops being enough.
Sample: AI workflow buildout
A build-log style post for showing the before/after shape of an AI-assisted operational workflow.
AI infrastructure
Retrieval, evals, observability, source boundaries, and infrastructure choices that matter when healthcare data is messy or sensitive.
No AI-infra case published yet.
Strategy and diligence
Build-vs-buy, vendor claims, operating leverage, adoption constraints, and the questions investors should ask before believing the workflow story.
Sample: investor diligence memo
A memo format for testing a company's AI workflow claims against source evidence and operational risk.