Hi! I'm

Harrison
Padgett

Computer Science Undergrad @ UMD

Harrison Padgett

Experience

Software Engineer Intern

Nokia
Jun 2026 — Present
  • Developed software for the Cloud Mobility Manager (CMM) team.

Machine Learning Researcher

University of Maryland (Dept. of Computer Science)
Jan 2026 — Present
  • Researched black-box LLM hallucination detection methods, evaluating generative AI systems across 4 established NLP benchmarks.
  • Built a hallucination detection pipeline using a weighted ensemble of 7B LLM judges (Phi-3-Mini, Mistral-7B, Qwen2.5).
  • Achieved 94.13% accuracy and AUROC 0.9212 across 9,960 TriviaQA questions.

Software Engineer Intern

Johns Hopkins Applied Physics Laboratory
May 2025 — Aug 2025
  • Built modular Python tools for large-scale data analysis of over 10 million signals.
  • Applied unsupervised clustering algorithms (HDBSCAN) to deinterleave and enhance signal separation accuracy.
  • Accelerated analyst workflow by over 90%, reducing task duration from 10-15 minutes to seconds.

Web Developer

University of Maryland (Adele H. Stamp Student Union)
Oct 2024 — Sep 2025
  • Engineered full-stack features for UMD's student center using C#, ASP.NET, MySQL, and Bootstrap.
  • Boosted performance and maintainability of systems serving over 40,000 students, staff, and faculty.
  • Upgraded the University Health Center mobile app with meal planning, food logging, and goal-setting features.

Tech Team Organizer

Bitcamp
Oct 2024 — Present
  • Developed Bitcamp’s website and mobile app, updating layouts, event information, and UI features using Nuxt.js (Vue 3), React Native, and TypeScript.
  • Contributed to a front-end subteam that built the organizers' roster page, ensuring a responsive design across platforms.
  • Supported the experience for 1,000+ participants by maintaining accurate, accessible, and up-to-date content across Bitcamp’s web and mobile platforms.

Undergraduate Researcher

First Year Innovation and Research Experience (FIRE)
Aug 2024 — Dec 2025
  • Processed large environmental datasets with R/ tidyverse to reveal trends using ggplot2.
  • Built Python XGBoost models to predict natural disaster injury outcomes, evaluating with RMSE/ R² metrics.
  • Presented statistical findings on AI vs. human customer support at a faculty-reviewed poster session.

Volunteering

Student Advisory Board Member

University of Maryland (Office of Undergraduate Research)
May 2026 — Present
  • Provided critical student perspectives to shape OUR's mission, vision, and values.
  • Contributed to discussions guiding campus-wide undergraduate research programs, services, and initiatives.

Peer Research Mentor

First Year Innovation and Research Experience (FIRE)
Jan 2026 — Present
  • Mentored incoming students in the Computing & Society stream, teaching R programming, dataset analysis, and qualitative research methods to explore the intersection of technology, equity, and social justice.
  • Facilitated stream operations as a liaison between students and the FIRE Faculty Leader, guiding peers through the entire research process from experimental design to IRB navigation and academic publication.

Projects

Poker AI

Reinforcement Learning / C++ / PyTorch / Deep CFR

Engineered a poker AI using Deep CFR to learn a provably unexploitable Nash equilibrium; re-implemented the game engine in C++ for a 35x speedup.

Play Demo GitHub

Occupan-See

PyTorch / React / Firebase

Supported development of an award-winning application that uses computer vision and Mask R-CNN to identify potentially dangerous occupancy levels in a room.

Devpost GitHub