Co-developed 17 specialized, single-purpose AI agents and contributed to a multi-agent orchestration flow using dynamic routing to dispatch tasks based on domain context.
Integrated a hybrid SQL/MongoDB persistence layer to store and retrieve execution context and prior state, including error logs and tmux sessions, improving data retrieval for recurring agent workflows.
Built an automated testing harness using LLM-as-a-Judge patterns to validate agent behavior and output integrity, reducing reliance on manual review.
Systems & Infrastructure Intern
University of Nevada, Las Vegas — Office of Information TechnologyHPC decommission
Supported decommissioning of a High-Performance Computing (HPC) cluster by documenting and auditing GPU compute servers and InfiniBand networking hardware to support secure asset recovery.
Audited fiber pathways across multiple facilities and synchronized physical assets with Schneider Electric EcoStruxure DCIM records to improve data center infrastructure accuracy.
Reviewed network/security logs with IT operations staff to identify anomalous SNMP authentication events; used DNS/MAC lookup tools to help trace source devices.
Network Engineer Intern
University of Nevada, Las Vegas — Office of Information Technology660+ switches
Collaborated with a 15-person network engineering team on a campus-wide Cisco-to-Juniper migration, including switch rollouts and wireless AP replacements.
Developed standardized configuration templates and automated IP management workflows that helped maintain consistent VLAN/port-level logic across assigned switch groups in a 660+ switch environment.
Assisted with provisioning and post-migration validation of network assets within a ~4,500-device campus portfolio, supporting connectivity testing during cutover windows.
Education
Bachelor of Science, Computer Science
University of Nevada, Las Vegas
Relevant coursework: Object Oriented Programming, Data Structures and Algorithms, Analysis of Algorithms, Intro to Machine Learning, Operating Systems, Cloud Computing.
Projects
Rebel Locate
Python, KNN and CNN models88% accuracy
Developed a geolocator and scene-labeling tool for UNLV, achieving 88% accuracy by performing data analysis on metadata from over 6,000 images to identify building proximity using a KNN algorithm and assigning scene labels (e.g., classroom, hallway) via a dynamically trained CNN model.
Skills
Languages
Java, C++, C#, C, Python, x86, MIPS
Frameworks
TensorFlow, Pandas, NumPy, LLM evaluation
Tools
AWS, PostgreSQL, MongoDB, Docker, Git and GitHub, InfoBlox, NetBox, SolarWinds, Schneider Electric EcoStruxure, Cisco IOS, Juniper Junos CLI, OpManager