At TenX, I build automated testing frameworks for enterprise multi-agent LLM systems, evaluating prompt quality, reasoning steps, agent handoffs, output drift, and edge cases.
As Transect's first data hire, I built and owned its BigQuery warehouse, dbt project, and Airflow pipelines across product, sales, and operations. I also built a retrieval pipeline over private permit and zoning documents and evaluation harnesses that track model accuracy and hallucination rates.
Previously at SSI and Jeeny, I designed streaming and batch data pipelines, production forecasting data models, semantic layers, and data validation systems processing hundreds of millions of daily events and supporting a 10M+ annual-trip marketplace.
I'm hands-on with Python, SQL, dbt, Airflow, BigQuery, AWS, Kafka, RAG, Claude, OpenAI APIs, and LLM evaluations, and I focus on taking systems from raw ingestion through dependable production delivery.
