CV

Applied Scientist II Intern (L5)

Amazon

Optimized large language model (LLM) performance for real-time applications.

• Delivered scalable knowledge distillation approach for internal LLM application, improving inference latency by 4.5x while maintaining accuracy
• Collaborated with cross-functional teams to deploy distilled models into production at scale
• Developed internal artifacts for vLLM deployments, Supervised Fine-Tuning (SFT), and Reinforcement Learning with Human Feedback (RLHF)

Machine Learning Engineer Intern

Northlight AI

Drove real-world AI solutions for underserved markets by deploying pragmatic, ethical machine learning models in cloud environments to empower small-to-midsize enterprises.

• Delivered a customer-facing product that resolved pain points for small businesses by leveraging LLMs and reinforcement learning to optimize supply-chain operations

Machine Learning Research Assistant

UNH RL² Lab

Studying machine learning and reinforcement learning (RL) under Dr. Marek Petrik in the RL² Lab.

• Doctoral research publishing new methods for risk-averse decision-making problems, with a focus on RL and LLMs
• Efficiently implementing and testing RL algorithms in Julia and Python (github.com/RiskAverseRL)
• Job skills: Python, Julia, TensorFlow, PyTorch, LLMs, Risk-Sensitive Decision Making

Data Scientist Intern

NextStep Health

• Implemented on-prem models to predict patient outcomes using Python and TensorFlow
• Delivered internal tools that reduced turnaround time for data analysis by 30%

Software Engineer Intern

Liberty Mutual

• Fostered code reuse by creating a shared Python library, reducing code duplication by more than 80%
• Optimized new data storage and retrieval system using AWS S3, reducing on-premise storage costs by 25%

Senior Test Engineer

UNH Interoperability Lab

• Reviewed and engineered test media for IEEE Standard 802.3, improving test coverage for back-plane and Base-T
• Job skills: Python, C++, Java, Test Engineering, Standardization, IEEE 802.3, Robotics, Hands-on Experience

Master of Science

University of New Hampshire

Computer Science, Summa Cum Laude (GPA 4.00 / 4.00)

Doctor of Philosophy (PhD)

University of New Hampshire

Computer Science (expected May 2027)

• Advanced Algorithms in Machine Learning
• Deep-Q Learning
• Reinforcement Learning
• Risk-Sensitive Decision-Making
• LLM-Driven Optimization
• Foundation Model Theory
• PyTorch, TensorFlow
• Prototyping, Implementation, Robotics

Bachelor of Science

University of New Hampshire

Computer Science, Summa Cum Laude (GPA 3.98 / 4.00)

• Data Structures and Algorithms (C++)
• Probability and Statistics in CS (Python)
• Intro to CS II (C++)
• Linear Algebra with Computational Applications (Python)
• Reinforcement Learning (Python)
• Neural Networks (Python)
• AWS, Azure, REST
• Software Development, Cloud Infrastructure

Presented at the first Reinforcement Learning Conference (RLC)

RLC 2024

Technical Reviewer

Reinforcement Learning Journal; European Workshop on Reinforcement Learning (EWRL)

Graduated Summa Cum Laude (B.S. and M.S.)

University of New Hampshire

Presented at the UNH Undergraduate Research Conference

University of New Hampshire

Dean's Award for Academic Excellence in Computer Science

University of New Hampshire

President's Scholarship

University of New Hampshire

Awarded for academic achievement and leadership, two years in a row.

Dean's Scholarship

University of New Hampshire

Scholarship awarded to students with exceptional academic performance, scholarly excellence, and leadership qualities.

Phillips Family Engineering Scholarship

University of New Hampshire

Awarded to Engineering students who maintain a 3.5 GPA or higher in the first 2 years of study.

Computing Monetary Risk Measures in Linear Time

Transactions on Machine Learning Research (TMLR)

Developed a linear-time algorithm for computing monetary risk measures, which are widely used in reinforcement learning and finance. Optimized state-of-the-art methods for computing risk measures, resulting in a log-factor speedup in computation time.

Risk-Averse Total-Reward Reinforcement Learning

Advances in Neural Information Processing Systems (NeurIPS)

Developed risk-sensitive reinforcement learning algorithms that optimize for the risk-adjusted future rewards, rather than the expected reward. Gave practical algorithms for optimizing the entropic utility of the total reward, which is a common risk-sensitive objective in domains such as finance, health and autonomous driving.

Deep Reinforcement Learning-Based Optimization of an Island Energy-Water Microgrid System

Resources, Conservation and Recycling (RCR)

Interdisciplinary work with the UNH Civil Engineering Dept. Developed a deep reinforcement learning model to optimize the energy and water usage of an off-shore microgrid. Real-world RL application which resulted in a 20% increase in island sustainability, with less human intervention.

LLM Thematic Analysis for Agriculture Research

Led collaboration with UNH Agriculture Dept. to develop a thematic analysis pipeline for categorizing research laboratories across the USA. Implemented a human-in-the-loop LLM system to improve accuracy and reduce workload, leveraging LangChain.

ROIL: Robust Offline Imitation Learning

Reinforcement Learning Journal (RLJ)

First publication. Led novel research on a robust method for training agents from data using offline imitation learning, surpassing previous work by leveraging consistent policy sets. Improved on existing IRL methods by allowing for covariate shift, which is common in the real world.

PPO: Proximal Policy Optimization

Open source project on a flexible implementation of PPO. Implemented in Python. Using PyTorch Lightning.

• Deep RL
• RL
• Python

Bandit Algorithms

Open source project on bandit algorithms. Implemented in Python. Following the popular book by Tor Lattimore and Csaba Szepesvári.

• Bandits
• RL
• Python