Multi-objective alignment of LLMs with the Rufus team.

My research focuses on improving the data efficiency of reinforcement learning (RL) via (1) adaptive action selection and (2) data augmentation. My goal is to make RL a more viable tool for real-world tasks. Towards this end, I’ve been using physical robotics tasks in my research.

- Fall 2021: TA for Mathematical Foundations of Machine Learning (CS 761)- Fall 2021: Head TA for Intro to Computer Systems (CS 354)- Spring 2021: TA for Problem Solving for Engineers (CS 310)- Fall 2020: TA for Discrete Mathematics (CS 240)

- Built the query execution and storage engines of Hustle, a scalable data platform built on top of Apache Arrow.- Designed a variant of the Lookahead Information Passing (LIP) query execution strategy with improved robustness in dynamic data environments for Hustle. [github]

- Deep reinforcement learning for power systems management. - From May 2022 - Nov 2023, I served as a consultant for reinforcement learning projects.- Received the 2023 Sandia Employee Recognition Award, which is awarded to < 10% of the Sandia workforce.

- Designed and optimized the first search for new hadronic WbJ states in Upsilon(5S) data collected by the Belle experiment.- Created tools to monitor TOP Level-1 trigger performance in the upgraded Belle-II Experiment.- Presented at the 2018 American Physical Society (APS) April Meeting.

- TA for Quantum Mechanics (PHYS 1370)

- Established genuine relationships with customers and developed a reputation for sincere service. - Improved customer satisfaction by effectively resolving issues and complaints. - Managed tobacco and lottery sales. - Assisted in front-end management.

Assisted customers with purchases and answered store related questions.