Cohort

Cohort pulls clinical and demographic data from your EHR over FHIR R4, normalizes it, and lands a tidy patient-level table in one warehouse so your analysts stop stitching exports together by hand. It refreshes on a schedule, keeps a record of what changed, and flags rows that fail validation rather than silently dropping them.

Each run resolves duplicates, maps conditions and encounters to a stable schema, and groups patients into the cohorts you define so risk stratification and care-gap reporting run against the same source every time. Definitions live in version control, so a cohort means the same thing in March that it did in January.

Expect first clean cohorts within a few weeks once source access is sorted, and refresh lag that usually sits in the minutes-to-low-hours range depending on volume. Data quality at the source caps accuracy more than the pipeline does, and Cohort surfaces those gaps instead of hiding them.