Austin, Texas, United States
I help scientists and researchers make real business impact through good engineering. I believe that relationships matter more than org charts, learning happens best through intelligent action, and that teamwork is leadership. Some things I'm proud of include the students I have helped learn to code, my pun generator, and a talk I delivered at one of the largest data science conferences. My interests include trustworthy models, imbalanced data, and (of course) putting machine learning into production. I studied computer science at Baylor University and have worked for companies including TripAdvisor and Indeed. I currently work as a Senior Machine Learning Engineer in Austin, TX.
Outbound lead scoring - Designed and implemented a machine learning system to rank leads for a sales team - Trained logistic regression (scikit-learn) and GBDT (CatBoost) models on Kubeflow. Wrote FastAPI application to deploy models in Google Kubernetes Engine with DataDog monitoring. Used Bigtable as feature store. - Set up and ran experiment showing 29% uplift in conversion rate Lookalike modeling - Used unsupervised learning techniques to recommend new prospective business customers to insurance agents - Created distributed training pipeline using Kubeflow and Spotify Annoy. Defined system interfaces and built initial embedding model (random projection). Productionized a contrastive learning approach built with TensorFlow. - Designed A/B test
- Used Spark, Hive, and Scala to gather, transform, and analyze data - Built, trained, and monitored machine learning models used in a large-scale production system - Wrote maintainable, scalable, production-level Java code used to serve large volumes of traffic - Wrote and reviewed design documents and system architecture documents in collaboration with Software Engineers
- Helped executive stakeholders answer business questions with data, including metric definition, analysis, and communication of results - Wrote high quality ETL code, including integration with external API’s - Provided technical leadership for a team of six developers to create a data warehouse - Designed and developed a web scraping system with Python, Postgres, and Tableau that is used by C-level executives - Created a customer segmentation model using PCA and kNN
- Develop a stream processing platform using Kafka and Python - Implement reliable ETL pipelines - Improve development and release processes through testing and automation
- Full-stack Java web development on the Search Engine Optimization team - Implement tests leading to gains of 60 million annual referrals - Enhance maintainability by removing outdated code and improving automated testing