Computer Science · Research

Alan Pysnack

M.S. Computer Science Student at UMBC

I am a Master's student in Computer Science at UMBC exploring how artificial intelligence can mediate and support human communication. Rather than using AI to replace human interaction, my research focuses on designing generative tools and interactive games to help people build social connections. Prior to graduate school, I worked for four years as a software engineer at Assurant. My background building web and mobile applications directly informs my current work in developing and deploying interactive research prototypes.

Email ORCID GitHub LinkedIn CV Experience ↓

Graduate Research

Current graduate research and project work at UMBC.

Graduate Research · 2026 In progress

Using AI-Generated Images to Facilitate Social Connection

This project explores how generative AI can help strangers move past surface-level small talk and build meaningful connections. By transforming a participant’s lived experiences into a personalized visual artifact, the game asks their partner to interpret the image—creating an indirect mechanism for reciprocal self-disclosure and deeper conversation. Beyond evaluating initial connection and enjoyment, future work aims to incorporate adaptive feedback loops, allowing the system to dynamically tune game parameters (such as humor or vulnerability) to optimize social outcomes.

Human-AI Interaction AI-Mediated Communication Generative AI Adaptive Systems
Graduate Research · 2026 In progress

Does Adding a Source Make LLM Answers More Trustworthy?

This project investigates how different types of evidence affect the factual reliability of large language model answers. Using questions from SQuAD 1.1, the study compares answers generated without evidence against answers generated with relevant, irrelevant, or conflicting information. The planned evaluation examines answer accuracy, whether answers are supported by the available evidence, and how models respond to contradictions.

Trustworthy AI Large Language Models Experimental Design Python

Experience & Past Projects

Professional software engineering background and undergraduate work.

Professional Experience · 2022–2026 Software Engineering

Associate Software Engineer at Assurant

Developed and maintained production web and mobile applications using Ruby on Rails, React, React Native, and TypeScript. Worked on REST and GraphQL APIs, automated testing, code reviews, and technical design. Improved pagination and client-side caching in a high-volume workflow, reducing API requests by approximately 2,000-fold.

Ruby on Rails React React Native TypeScript API Design Testing
Senior Capstone · 2021 Computer Security

Addressing PCR Fragility in TPM 2.0 Systems

Investigated how changes to a platform's configuration, such as firmware updates, can invalidate TPM 2.0 policies protecting sealed encryption keys. Studied policy mechanisms in the TPM Software Stack and implemented a Policy OR approach in Kmyth, an open-source TPM 2.0 security framework maintained by the National Security Agency, extending sealing functionality to accommodate multiple expected platform states.

Computer Security C TPM 2.0 Open Source
Undergraduate Research · 2021 Machine Learning

Classifying Protein Signal Peptides

Evaluated supervised machine-learning algorithms for predicting signal peptides in eukaryotic and archaeal protein sequences. Prepared datasets, trained and tested classifiers using Weka, and compared their performance using Matthews correlation coefficient (MCC).

Machine Learning Classification Weka Bioinformatics