Integrating SAS solutions with modern lakehouses for medication adherence risk often slows teams down because mapping source data into a required common data model can take weeks of manual engineering. In this session, we'll demonstrate a low‑code approach that uses SAS access to DuckDB plus S3 table buckets (Iceberg) to land raw healthcare data cheaply, then uses a “Data Mapper” agent to map source metadata to a target SAS schema without touching the underlying data. You’ll see how the agent generates mappings, flags items needing human review, and produces executable SQL that creates views or tables shaped exactly for the SAS Medication Adherence Risk pipeline. The result is a faster path from messy source files to governed, solution‑ready data structures, with human-in-the-loop transparency built in.
Presenters: Joe Cabral and Jim Georges, SAS
Watch the recording
