Run statistical studies on market data in Python. Research applications of modern NLP techniques in predicting market information. Research non-standard methods for modeling market impact.

- Develop bidding algorithms and infrastructure for real-time bidding in online auctions for online advertising that operate on the scaleof hundreds of milliseconds- Create a counterfactual simulator for forecasting auction supply and demand dynamics using Spark via Pyspark with AWS EMR, processing tens of billions of datapoints

- Develop a cloud-based solution for solving the “last mile” problem in a logistics network, using microservices and Java Spring in AWS with DynamoDB- Led a phased migration of configuration data from a legacy system into DynamoDB, including the addition of a REST API and scripts to automate managing configuration data- Led a data mining effort using Snowflake, SQL, and Python, analyzing data to predict optimal times for delivery success

- Work at University of Washingotn's "Noah's ARK" Natural Language Processing research group- Researched state of the art Natural Language Processing machine learning techniques, focusing on deep learning with transformersand neural networks- Created a corpus for text-based content-aware recommender systems, using Reddit data to examine user behavior over time- Experimented with novel techniques for automatic domain resolution and adaptation for improving zero-shot performance- Developed and trained a neural model to discover relations between scientific papers and analyze citations- Investigated techniques to enhance interpretability of the attention weights in natural language models