Ovian Thomas

I am a senior AI/ML engineer with 9+ years building production systems across backend engineering, healthcare ML, MLOps, and large language model products. I focus on making LLM systems dependable in real-world production by emphasizing reliability, observability, and safety.

At Anthropic I designed and deployed production LLM systems, agentic workflows, and RAG pipelines, improving latency, cost efficiency, and task completion for customer-facing experiences. I led eval and observability pipelines and mentored engineers on robust AI system design and deployment.

At AWS I owned MLOps pipelines, model registry-driven releases, containerized inference services, and IaC/CI-CD automation to improve repeatability and deployment governance. My work reduced inference processing time and increased release confidence through integrated model quality and bias checks.

Earlier roles include oncology-focused ML and NLP at Flatiron Health and backend/cloud modernization at Accenture, where I built predictive models, ETL pipelines, REST APIs, and migration playbooks. I prioritize practical engineering decisions that improve grounding, latency, and reliability for production AI systems.