- Product Owner for Liberty Mutual's enterprise data science tools and platforms that span the entire machine learning lifecycle, from data prep and model training to model deployment and monitoring

- Product manager of in-house machine learning (ML) model deployment platform within Liberty Mutual's Office of Data Science- Work closely with, and bring together, teams across data science/advanced analytics, engineering, product, and the business to figure out why, who, what, and when- Define product vision and roadmap to stay ahead of the curve in the model deployment and other related aspects of MLOps- Work with other internal platform teams to provide a more seamless and cohesive data science model lifecycle management experience to data scientists and related stakeholders- Consult with data science teams to speed up deployment and operationalization of ML models- Evangelize and strategize with a wide range of technical and non-technical stakeholders and executive leadership- Uphold the traditional responsibilities of a product owner on a scrum team

- Managed and implemented new features in our generalized model deployment solution to capture model inputs and predictions for downstream analytics- Developed the pure-Python back-end API and ETL process for a customer-facing application that helps our underwriters triage and prioritize submissions from brokers using predictive models deployed on our in-house model deployment platform. (Python API/ETL, AWS RDS, Docker)- Contributed to an in-house generalized machine learning model deployment framework (Python, Bamboo CI/CD, Docker)- Acting Product Owner for new in-house model deployment platform

- Aimed to solve the “model deployment dilemma” in Liberty Mutual by creating a self-service model deployment framework for any data science team that generalizes and expedites the current deployment process to reduce model backlog and mitigate the all-too-common “$1 million deployment” problem - Contributed to the platform's Flask-based Docker layered service architecture- Largely responsible for product communication (PowerPoint presentations, documentation, product branding) to internal stakeholders, customers and peers- Completed various exploratory data science projects using Python, including developing a machine learning model to predict the salvage value of a vehicle at First Notice of Loss based on vehicle make, model and model year using internal claims data

- Planned and executed an 8-week internal data analytics competition (“hackathon”) for roughly 200 participants focused on AWS machine learning tools to spur interest in cloud computing and machine learning and explore business opportunities in the workers’ compensation space (resulting in at least two go-forward ideas)- Researched and identified the ideal toolset and large dataset for the hackathon among several vendors, cloud provider alternatives and countless datasets by assessing first-time use friction, internal interest, pricing feasibility and multi-team scalability- Developed and provided basic demos and instructions to participants on the use of AWS tools and predictive modeling practices

- Created a new archival system of electronic and physical design history files to meet internal and FDA regulations- Analyzed global sales and specification data of released products in Excel to determine most critical feature sets for new product in development- Communicated new organizational system to leadership team, including presenting new system and conducting meetings- Developed, managed, and implemented process to digitize engineering design history files

- Led and managed small teams of three to four high school interns and assigned tasks as needed- Planned and executed projects, such as computer lab design and construction, throughout the school district- Consulted with district staff on purchase orders, bills of material, and project design