The canonical 5-layer architectural specification connecting business decisions, bitemporal data foundations (SCD2 & dbt integrity checks), enterprise BAA model gateways, human-in-the-loop agent orchestration, and operating cadence.
Insights
Healthcare data strategy, dbt, domain knowledge, and the parts of analytics work that must survive real operating pressure.
Latest articles & specifications
Start here if you want the thinking behind the consulting work: domain fluency, model trust, and analytics systems that business teams can act on.
Why treating healthcare claims like e-commerce receipts corrupts financial ratios. How to model line splitting, CAS adjustment segments, overpayment takebacks, and true net denial rates in dbt.
Read the guideHow we audit a healthcare dbt repository in 48 hours: diagnosing surrogate key grain, bitemporal enrollment, midnight timestamp precision drops, NPI/ICD validity, and referential integrity using open-source checks.
Read the playbookClaims are not real-time event streams. Why naive AI pipelines hallucinate retrospective facts, what every timestamp actually encodes, and the strict engineering guardrails required for healthcare data.
Read the articleA practical framework for choosing a single-provider path or orchestrating multiple models across labs and providers.
Read the articleA practical way to connect business decisions, data products, model strategy, workflow automation, governance, measurement, and adoption.
Read the articleA feature on Reconfigured highlights why technical skills alone aren’t enough in regulated data environments — and how deep domain knowledge prevents costly mistakes.
Read the articleHow dbt keeps our healthcare data platforms transparent, documented, and ready to scale with every new question.
Read the article