Projects across FHIR, claims, EHR, and quality — enabling interoperable data platforms, integrity monitoring, and care performance.
End-to-end ingestion of FHIR resources (Patient, Encounter, Observation, Claim) into curated Snowflake models. Includes flattening, schema validation, ICD-10 / LOINC enrichment, SCD2 dimensions, and dbt tests for data quality.
Claims profiling and anomaly scoring using ICD-10 groupers, utilization patterns, cost outliers, and frequency spikes. Generates prioritized review queues and summary dashboards for integrity teams (FWA).
Combines EHR encounters, claims, and provider data into marts for utilization, length of stay (LOS), readmissions, and throughput. Exposes flexible filters for payer type, diagnosis groups, and service lines.
Models patient flows across in-network and out-of-network providers to quantify leakage, referral patterns, and high-opportunity service lines. Supports network optimization and contracting strategy.
Curates patient-level features like comorbidities, prior utilization, SDoH proxies, and discharge disposition into an ML-ready feature store for 30-day readmission and care management prioritization models.