Experienced in scaling platforms, projects & infra for hyper growth. In the past I've created instrumental portions of our Fraud prevention infra, from data & feature engineering to higher level business domains.

Currently building out the Trust Platform to support a multi-product line future.

ML Ops: Application, data, and some infrastructure engineering. Productionized machine learning models for scalable API use, developing a pricing rules engine, and trying to construct relationships between their outputs.

Product Engineering: Built out features around carrier tracking, automated communication (customer notifications), and other features for the company's internal load board.

Languages: Python, Ruby, Postgres, Ruby (Rails), Java, Node (a little)

Tools: Lambda, EC2, ECS, SQS, SNS, Elasticache (Redis), RDS, ECS, Redshift, Snowflake, Redis, Airflow, Terraform, CircleCI, etc.

Applied and evaluated physical simulation regressors to data in a distributed-memory parallel context (LLNL's supercomputing infrastructure). Researched current learning methods on imbalanced class distributions and used those to composed locally trained classifiers for a node's training subset. Reduced those models into a superset classification algorithm while retaining performant regression versus our baseline comparison.

In layman's terms, find an untuned model, train an instance on each compute node's portion of the data, combine all nodes' trained models in some fashion (the hard part). Choose models based on research & intuition, tune, and integrate the resulting model into existing experiments.

Tech used: Python, scikit, MPI

Voodoo Manufacturing was a Y-Combinator company focused on creating a fully automated factory of 3D printers, to enabling rapid prototyping and low-scale manufacturing.

In my internship, I designed the first iteration of an automated job assigner to manage factory load. I also built out parts of the client-facing website and internal tools. Learned a lot about choreographed event driven architectures.

Tech: Go, Rails, React/Node, Kafka

HIV research underneath Nicole Frahm and Bryan Simons.