- Built an NFL schedule optimization engine which helps business partners determine which games are expected to bring in the highest TV ratings at both the local market and national levels and then assigns each local market a specific game from window’s lineup.- Developed real-time insight recommendation systems for both NFL and MLB live games. System used current game state and player information to rank contextually relevant content to be used as a graphic or enhance the overall narrative of the game.- Created an NFL game simulation framework using a series of well-calibrated probabilistic models. Built a Streamlit app to serve as a front-end for business users to interact with the models and explore different scenarios. The results of the simulation framework were used in several applications - insight recommendation engine, win probabilities, 4th down decision model, and others.

- Machine Learning & Intelligent Operations Garage- Developed intelligent call routing strategy for card collections. Optimally matched customers with specialists based on attributes that were predictive of net credit loss reductions. Used Python, Spark, and Hive for strategy implementation.

- Built an attribution model to quantify the influence of different components on the change in value of derivatives over time. Exposed model functionality with Flask/React.js app to allow consultants to run models on specific portfolios and share reports + data visualizations with clients.- Engineered data pipeline to reduce vendor API expenses by scraping natural gas prices from public pdfs published on an exchange’s website. Automated with Selenium and Python.- Created a Gitlab-CI pipeline framework to allow other analysts/data scientists to automatically deploy new features to internal data tools effortlessly, saving several hours of manual work a week.- Multi-session speaker at company sponsored "Python Boot Camp for Actuaries" conference (2018, 2019, 2020). Gave talks on building Flask apps, web scraping, and Pandas+Numpy libraries.

- Cross channel analytics for card acquisition strategies and competitor/market research.- Worked with large datasets (billions of records) daily, including structured data (consumer credit profile, spend behavior, ...) and unstructured data (web visitation, search, ...).- Presented new strategies to senior leaders regularly.- Used technologies such as Teradata, PostgreSQL, SAS, KnowledgeSEEKER.

1st Rotation: Branded Card Portfolio Analytics2nd Rotation: CCAR Stress Testing & Forecasting3rd Rotation: Acquisition Marketing Analytics