At Oracle, I design and ship RAG pipelines that connect internal knowledge sources to LLM-backed services, improving answer relevance for enterprise users. I also built evaluation frameworks to measure accuracy, hallucination rates, and regressions across model versions.
I architect AI platform components in Python, standardizing how teams integrate vector search, embeddings, and prompt orchestration with LangChain and LlamaIndex. I established safety standards and technical review gates for production AI features.
At Blue Origin, I built backend platform services in Python and event-driven data pipelines with Kafka. I also developed machine learning components with PyTorch and TensorFlow to support model training and batch inference.
At Nordstrom, I built features for customer-facing retail applications and implemented REST APIs supporting product catalog and checkout flows. My work has also included distributed systems, cloud deployments, observability, and CI/CD.
