David Starr Email & Phone Number

New York, New York, US

Carpinteria, CA, US

Continuing to build out our ML platform. Delivered our first ML models on top of Kubernetes in production (using mlflow, later Seldon).

Carpinteria, CA, US

Began building our ML and data platform. Led development starting from zero data access to a secure environment for model training and feature development, supporting data processing pipelines in Airflow and large-scale compute with ephemeral Spark clusters. Stakeholders included internal stakeholders (IT, revenue) as well as traditional user-facing.

Partnered with product to research and prioritize the data roadmap. Built containerized pipelines to clean and enrich our database and train and evaluate ML models using airflow and mlflow. Built functioning prototypes of candidate features and estimated their feasibility. Created a menu of data collection roadmap options along with their corresponding.

Engineering team co-head during a year of growth between CTOs. Restructured the development organization into self-sufficient & cross-disciplinary squads. Grew the engineering team from 14 to 30, including a healthy mix of on-site and remote, ranging from fresh grads to 10 and 20 years experience, as well as our first dedicated engineering managers.

Tech lead responsible for one of two engineering teams. Oversaw a self-sufficient group of developers who managed and executed our work across the entire stack: back end (Dropwizard/Java), front end (Angular/JS), and QA. Worked closely with the product and design teams to refine the product roadmap.

Washington, District Of Columbia, US

Architected and built our 100 GB - 1 TB -scale (batch) data processing and machine learning pipeline in AWS using Celery, Luigi, and Apache Spark.Developed machine learning models to predict consumers’ habits and brand preferences, including the models which powered our flagship product for our biggest client. Training data ranged in size from 10k to 10mm.

New York, New York, US

Created statistical models to forecast the supply, demand, and price of various commodities (electricity, renewable energy, and natural gas) from underlying driving factors and interpreted those models to identify trading opportunities.Co-designed and implemented a flexible and modular infrastructure to encode, interpret, and value (often bespoke).

Phd, Physics

Sb, Mathematics, Physics