Healthcare technology
Connecting software development with healthcare production
Production performance leadership and a controlled maintenance process for a demanding healthcare platform, including 150GB+ indexes.
Steven Wittek’s prior employment and consulting experience · Organization anonymized
The work
Healthcare benefits software sits behind operational work that health plans and pharmacy benefit managers depend on. When production workloads slow down, the issue reaches beyond a database dashboard: application processing and the people relying on its output are affected. Changes also have to respect restrictions on protected health information.
I served as a senior U.S.-based database performance engineer and a technical liaison between offshore software developers and production operations. Differences between development and production workloads, together with restrictions on production data access, made it difficult to investigate some problems entirely within the development environment. I connected those teams through production-aware technical investigation and review.
I reviewed T-SQL, stored procedures, execution plans, and indexing changes against Query Store history and actual workload behavior. My role was to turn symptoms such as high CPU into findings developers could act on, then evaluate the changes against the conditions that mattered in production.
Targeted query and indexing improvements helped keep production CPU below its previous roughly 80% level as demand returned toward seasonal peaks.
I also addressed an inherited index and statistics maintenance backlog extending beyond a year. With production indexes of 150 GB and larger, maintenance needed to be planned around the live workload. I established recurring maintenance and used controlled online, resumable operations for large indexes, accounting for transaction-log growth, concurrency, replica synchronization, and available capacity.
The work combined immediate performance investigation with a repeatable operational process. It gave development teams a stronger connection to production behavior and addressed maintenance work that had become difficult to complete through ordinary scheduling.
Engineering outcome
Improved healthcare production performance and established controlled maintenance for large database objects, while bridging distributed development and restricted production operations.