I coordinated the move of a legacy diagnostics data pipeline into cloud storage, with governed access for the data scientists who depended on it.
Abbott · Architect and Alinity Data Ecosystem
Challenge
Data access was slow and expensive to operate across a globally distributed diagnostics estate of roughly 27,000 instruments. The migration could not trade continuity for speed: the pipeline was in daily use, the teams depending on it were spread across time zones, and a cutover that lost data or access would stop work rather than merely slow it.
Action
Delivery ran across several systems at once — an Oracle-to-AWS S3 migration for the data itself, with Azure and SharePoint components carrying the data-science portal — and the dependencies between them, held by remote teams, were what made it difficult rather than any single move. My work was the integrated plan: sequencing the migration against portal readiness, naming who owned each dependency, and keeping the critical path honest as the scale of the estate surfaced. The Integrated Delivery Plan and Dependency Map held that sequence. The critical path was accepted, the cutover window approved, and rollback owned before anything moved.
Result
Data-retrieval efficiency increased by 65%, and a delivery cycle that had taken approximately six months came down to three weeks. The improvement came from the governed portal as much as from the storage migration: data scientists reached approved data directly instead of requesting it and waiting for someone else to retrieve it. The program recorded no FDA inspection findings, which is the evidence that the controls held under inspection rather than only on paper.
