At ONTAB, I develop transaction-oriented services and Python RAG pipelines, improving peak transaction response latency by approximately 32% and contextual retrieval accuracy by 32%. I also connect OpenAI and Claude services to document workflows, reducing manual review effort by approximately 30%.
I manage production workloads across AWS and Azure, maintaining approximately 99.95% availability for revenue-critical workflows. Previously, at BioMark and DEVSiNC, I developed full-stack applications, AI/ML pipelines, and product features for SaaS, fintech, healthcare, and other platforms.

