At Salesforce, I own backend and integration services for a virtual care platform, applying production AI/LLM patterns, structured data contracts, and reliable asynchronous workflows.
I rewrote a performance-sensitive Python service in Go, completing the same benchmark workload 27% faster and reducing memory use 83% while maintaining over 90% test coverage. I verified the replacement with benchmark parity tests, contract tests, shadow traffic, and staged cutover.
At USAA, I led delivery of a marketing decisioning engine and built NLG, ETL, and integration workflows for AI-enabled content and decisioning. I also modernized APIs through CI/CD and deployment-platform migration, with compatibility adapters and controlled cutover.

