CV

Degree complete

Ph.D. in Computer Science

Expected spring 2028.

Graduate coursework in Computer Science

B.S. in Computer Science, Summa Cum Laude

Graduate Research Aide

• Built a Slurm/Lustre testbed on CloudLab with fault injection across ~30 storage, network, and node-level fault types to evaluate LLM agents for autonomous HPC storage monitoring
• Designed multi-agent monitoring workflows and a grading framework that scores impact detection, root-cause attribution, and incident escalation against injected ground truth

W.J. Cody Associate

• Created a small evaluation suite of telemetry data analysis tasks for LLM agents derived from real tasks extracted from existing published works
• Developed an internal web service for ANL staff to submit telemetry analysis tasks in natural language; the agent executed SQL and Python workflows over 2+ years of Polaris production data and returned complete Jupyter notebooks with results.

Graduate Research Assistant

• Developed an LLM agent for diagnosing I/O performance bottlenecks by analyzing HPC application I/O telemetry
• Developed an autonomous tuning engine for parallel file systems by designing a multi-agent closed-loop system enabling rapid improvement of parameter settings
• Built a benchmark of realistic telemetry analysis tasks spanning over two years' worth of Polaris system data and used it to systematically evaluate LLM agents

Graduate Research Assistant

• Investigated the use of LLMs for anomaly detection in parallel file system logs
• Applied traditional ML-based classification techniques to predict severity of I/O interference among concurrent applications

Summer Research Intern

• Worked on improving dataset quality for domain-specific chatbots (pre-ChatGPT)

Undergraduate Research Assistant

• Worked on anomaly detection in parallel file system logs using semantic embedding, high-dimensional clustering, and bi-directional LSTM sequence classification models

Understanding Reliability Bottlenecks in LLM Agents for HPC I/O Telemetry Analysis

STELLAR: Storage Tuning Engine Leveraging LLM Autonomous Reasoning for High Performance Parallel File Systems

Augmenting LLMs for HPC I/O Performance Diagnosis

IOAgent: Democratizing Trustworthy HPC I/O Performance Diagnosis Capability via LLMs

Understanding and Predicting Cross-Application I/O Interference in HPC Storage Systems

ION: Navigating the HPC I/O Optimization Journey using Large Language Models

Early Exploration of Using ChatGPT for Log-based Anomaly Detection on Parallel File Systems Logs

Drill: Log-based Anomaly Detection for Large-scale Storage Systems Using Source Code Analysis

ClusterLog: Clustering Logs for Effective Log-based Anomaly Detection

Future Leaders of AI Doctoral Consortium

Selected as 1 of 53 (1 of 7 in Systems & Networking) invited participants for the Future Leaders of AI Doctoral Consortium at the 2026 ACM AI Leadership Summit.

Computer & Information Sciences Outstanding Graduate Research Assistant

Merit-based award given to 1-2 students annually in the CIS department

Department Travel Award for Outstanding Conference Publications

Awarded to students with first-author papers at top conferences

Winner of the Scalable Storage Competition (SSC24/25) hosted by GWDG (Individual track)

Awarded annually to one individual in the individual track of the competition

Mathematics and Computer Science Oral Presentation Award

Awarded to the top student presenter at the UNC Charlotte Undergraduate Research Conference

Software Developer

• Built supply-demand gap forecasting models for Charlotte-area employers (e.g. Atrium Health, Novant Health, Truist, etc.), integrating public labor market data at regional, state, and national levels with private employer data
• Designed and implemented a PostgreSQL database to securely store HRIS and ATS data from Charlotte-area employers
• Designed and implemented a high-performance Next.js backend to serve a data-visualization web app presented to members of the Charlotte Executive Leadership Council

Backend Developer

• Implemented and maintained test automation modules for IBM Cloud VMware services
• Improved load times of an internal test automation observability platform by 3x through backend refactoring
• Redesigned automation job configuration workflows to improve usability and flexibility