Held weekly consulting hours and graded assignments for CS260: Computational Genomics, taught by Dr. Alexander Hartemink.

Interned as a Software Engineer on the Data/Machine Learning (ML) Platforms Team at Convoy. Built a Model-Agnostic Machine Learning Performance Analysis Tool to help: (1) Improve the accuracy of the pre-existing machine learning models used throughout Convoy and (2) Streamline the model iteration process. Built the service using NodeJS, Typescript, React, Python, and AWS.

Worked on two projects related to incorporating machine learning algorithms and other novel algorithms into the backend infrastructure of AdStage's Automate product:

(1) Predictive Cross-Platform Ad Performance System: Using proprietary cross-network advertising data of AdStage's customers, I built a predictive modeling system to help customers gain insight into future performance of their advertising data across all platforms (Facebook, AdWords, Bing, LinkedIn, and Twitter). This modeling system was built using PostgreSQL to interface with AdStage's databases and Keras and Tensorflow to develop the machine learning models. The model was built using a stacked LSTM architecture and transfer learning.

(2) Budget Reallocation Optimization System: Using proprietary network advertising and budgeting data of AdStage's customers, I am currently building a budget optimization system to help paid lead-gen marketers automate the reallocation of their marketing budget across a selection of entities, ranging from Ads, Ad Groups, and Accounts. The budget optimization algorithm utilizes performance metrics and spend data to assist marketers in spending as close to their budget as possible.

In addition, I have assisted with various backend and frontend bug fixes.