At Amazon, I architect and productionize Python-based ML and data services on AWS, cutting latency by 20% across high-throughput workloads. I also lead LLM and retrieval workflows that reduce manual reporting effort by 35%.
Previously at sticky.io, I built data and ML services for customer analytics, recommendations, and AI-enabled SaaS workflows serving 50,000+ monthly active users. My work improved retrieval relevance by 25%, reduced analytical processing latency by 20%, and lowered API response times by 30% through database and caching optimization.
Across Amazon, sticky.io, and GrammaTech, I've taken models from exploration through deployment, monitoring, and decision support. I build with Python, SQL, Spark, Kafka, Airflow, cloud infrastructure, and MLOps tooling, while mentoring engineers on experimentation, ML infrastructure, evaluation, and reliable production practices.
