Notifications team - Responsible for all notifications to drive customer engagement & growth for DoorDash.

• Trained hybrid LightFM recommendation model with user web click and purchase history to improve email click-through by 22.9%

• Performed data wrangling and built a data pipeline using Snowflake SQL and Pandas

• Developed Gradient Boosting Regressor model for forecasting orders and reduced error by 20%

• Optimized hyperparameters using Bayesian Optimization to increase R2 value to 0.99

• Deployed fully automated recommendation model in Docker and Redis integrated with AWS S3

• Generated statistically significant insights from survey responses of computer science students to identify bias in algorithms and search engine results

• Presented overview and insights for two conferences - ALA Midwinter Seattle and code4lib San Jose