Andrew Ray

Designed a shared AWS ML workflow linking MLflow tracking and model registry with governed promotion. Standardized orchestration with Apache Airflow and AWS Step Functions, and provisioned training workloads with Terraform and Kubernetes.

Implemented Python training and batch-inference pipelines on AWS, integrating MLflow tracking and model registry. Built FastAPI real-time inference adapters and automated environment provisioning with Terraform and AWS CloudFormation.

Developed reusable Python feature and validation components for enterprise consulting engagements. Reduced feature-preparation runtime by 24% by partitioning Apache Spark transformations and reusing intermediate joins.

Integrated Python model components into software-services delivery, defining input contracts and REST interfaces. Shortened inference integration cycles by vectorizing feature transforms and preloading model artifacts.

Prepared insurance policy and claims data with Python and SQL, validating category mappings and missing-value rules. Reduced manual reconciliation by 2 hours per model refresh by replacing duplicated SQL extracts with reusable joins.

Supported data science tasks as an intern at Globe Life.

Master's degree of Science in Computer Science from Rice University, completed in 2017.

Bachelor's degree of Science in Computer Science from Rice University, completed in 2015.