Architected and implemented the MVP version of a flexible yet powerful machine learning platform to support applied machine learning features (including diverse use cases like NLP, binary classification, etc.) at Qualtrics using a system of AWS SageMaker and custom Dockerized components. Also defined the technical north star for the future of the machine learning platform.
Built features to supplement an intelligent, intuitive, and delightful statistical analysis software called Stats iQ using Python (numpy, pandas, TensorFlow) and JavaScript (NodeJS, React, Redux) backed by various datastores (Redis, MySQL) with a team of five other engineers. Some highlights: deprecating and rewriting a crosstabulation (Crosstabs) product, localization, a variable creation workflow (including variable bucketing and formula processing).
Designed, wrote, and maintained an operational tool used to monitor background job failures and other operational status indicators across multiple data centers in a single frontend interface written in JavaScript with React and Node, with data replicated in MongoDB clusters per data center.