Tristin Beckman

Tristin Beckman is a Senior Applied Scientist at Pendulum, where they have worked for the past six months. Prior to their current position, they worked as a Staff Data Scientist, Senior Data Scientist, and Data Scientist at Eventbrite for three years and eleven months. While at Eventbrite, Tristin developed a pipeline to get ML models to production using AWS tools such as SageMaker endpoints, DynamoDB, Lambda and API Gateway. Tristin also deployed this infrastructure with Terraform and built tools for the data science team to deploy batch production jobs. This included building a client for SageMaker to interact with presto/hive, as well as base classes to run jobs using Apache Airflow. In addition, Tristin developed classifiers from text data using Sentence BERT and Torch to support SEO teams understand what drives traffic to the platform. Tristin also used NLP tools to develop an event series detection model and built models to predict the financial impact of COVID-19 on the business. Furthermore, Tristin built a generalized forecasting tool to project revenue, tickets sold and creator acquisition across various markets. Tristin also analyzed the dynamics of event creators over time and studied the effects of pricing on event organizers and their attendees. Finally, they worked with a team to build a system to recommend ticket prices to event creators. Prior to their work at Eventbrite, Tristin was a Post-Doctoral Research Fellow in Political Economy and Experimental Research at the University of Oxford, where they developed novel measures of fiscal policy manipulation, built and managed databases, ran statistical analyses, presented results to expert audiences, and published research in a leading peer-reviewed journal (Journal of Politics). Tristin also worked as a Survey Administrator at EWALD & WASSERMAN RESEARCH CONSULTANTS, LLC for less than a year, and as a Consultant at Flowers Brothers Dominicana for four months.