At Snowflake, I built financial-intelligence enrichment workflows using Snowpark, Snowflake Cortex, Python orchestration, retrieval, and RAG. I integrated GPT, Claude, Llama, and Mistral for automated processing of external financial and news data.
I developed checks for source relevance, context completeness, and grounding, along with structured results and fallback handling. I also created repeatable hard-case evaluations to assess relevance, completeness, grounding, usefulness, consistency, and fallback behavior.
At Liberty Energy, I built and productionized AI/ML systems, Python services, APIs, and predictive-maintenance capabilities. I migrated analytics and ML workloads to Databricks and Apache Spark, and deployed a fuel-efficiency model that achieved 96%+ measured evaluation accuracy and contributed to millions of dollars in annual savings.
At ScienceSoft, I built backend and AI-enabled services for medical-imaging and bioinformatics workflows. My work included FastAPI services for routing DICOM images and a GPT-based bioinformatics pipeline using Biopython, AWS, and GenBank.
