The first enterprise data platform a Medicaid managed-care organization ran outside its core warehouse.

Everything the organization knew about its claims lived inside the core MMIS enterprise data warehouse, on that system's schedule and in that system's shape. Financial reporting and the verification of vendor remittance advice were largely manual, which meant slow month-end closes and defects that only surfaced when the numbers did not reconcile.

Underneath sat stored procedures of ten to twenty thousand lines with no documentation and no one left who had written them. Every change was a risk, and every new engineer took months to become useful.

I solutioned the organization's first enterprise data platform outside the MMIS warehouse: its own data governance, pipeline orchestration, and analytics capabilities, built so it could grow without waiting on the core system. Then I led architecture and development strategy for a 12-person team that automated the financial reporting and the remittance-advice verification on top of it.

For the undocumented code, I brought AI-assisted documentation into the SDLC: large language models produced first-draft documentation of the stored procedures, engineers verified and corrected it, and the result became part of the definition of done rather than a one-time cleanup.

In parallel I drove the partner-level architecture for modernizing EDI claims and encounters processing: the integration patterns, the data standards, and a platform design that scales beyond one line of business.