Writing

Writing

Healthcare systems architecture and data science are separate sections. Article addresses stay where they are.

Healthcare Systems Architecture

Payer business architecture, claims platforms, payment integrity, and the integrations behind healthcare operations.

Nov 2, 202612 min read
Healthcare payer business architecture: the five core entities
Part 1 of 6. Carrier, sponsor, policy, plan, and benefit. The same words point to different organizations in commercial coverage, Medicare, and Medicaid.
Sep 7, 20268 min read
Beyond pay-and-chase: a closed-loop payment integrity architecture
Pre-pay and post-pay only work together—fast inline scoring for adjudication, deeper retrospective analytics as the feedback loop that keeps the rules honest.
Aug 12, 20266 min read
Where payment integrity architectures usually break
Post-pay recovery fails less from missing rules and more from latency, brittle vendor seams, and claims data that no longer matches adjudication reality.

The data science lifecycle, data collection, machine learning, and the engineering that makes a dataset usable.

Nov 2, 202621 min read
Data collection and ingestion: building a reliable data foundation
Part 2 of 10. Build a reliable data foundation with source discovery, SQL and API extraction, batch and streaming patterns, data contracts, validation, orchestration, and lineage.
Oct 26, 202614 min read
Problem framing: the question before the model
Part 1 of 10. Translate a business problem into an analytical question, define the target and success metrics, and make the constraints explicit before choosing a model.
Oct 11, 20267 min read
Sampling variability: why two samples from the same population disagree
Sampling variability is the natural gap between statistics from different random samples of the same population. The standard error measures how far a sample mean can wobble, and the law of large numbers is why that wobble shrinks as the sample grows.