I'm rebuilding cloud-native data and document-processing platforms at Clara Analytics, where I replaced legacy Lambda and SQL workflows with Python microservices on Kubernetes. My work delivered 90% test coverage, a 4x reduction in incident diagnosis time, 37x throughput improvement, and 68% lower costs.
I've partnered directly with data scientists at Opensignal and Verisk to design ETL pipelines, productionize and retrain ML models, and build MLOps workflows with Databricks and MLflow. I work deeply across Python, Spark, Polars, AWS, Terraform, Airflow, and Kubernetes.
Earlier, I trained computer-vision models and developed the open-source cvdata library, while at NOAA I built climate-data processing systems and software for the U.S. Drought Portal.
