At MathCo, I developed a Python automation system that detects, diagnoses, and heals syntax and import errors in production pipeline files, with a 98% success rate. I also added observability for LLM token use, inference cost, and fixes.
Earlier at MathCo, I engineered a metadata-driven ingestion framework that reduced data latency by 60% and implemented automated validation checks that cut downstream reporting errors by 70%. As a Data Engineer Intern, I deployed a FastAPI service to Azure App Services and built a Medallion architecture on Databricks.
In personal projects, I built a Python Deep Packet Inspection engine for analyzing network traffic and a custom vector database and RAG system for context-aware semantic search. I work with Python, PySpark, Azure, and Databricks across data engineering and cloud projects.
