Gallery
Highlights from conferences, workshops, and research events
NASA MSGC Fall 2025 Conference
Our SAiRL (Security & AI Research Lab) team presented research on NASA-funded software vulnerability detection projects at the Michigan Space Grant Consortium (MSGC) Fall 2025 Conference. The presentations included oral talks and poster sessions showcasing our work on neuro-symbolic frameworks for spacecraft software security.
Oral Presentation
Oral presentation on spacecraft vulnerability detection
Conference audience during the presentation
Presentation Video
Poster Session
Student presenting the ResVul-LLM poster
ResVul-LLM: A Neurosymbolic Framework Combining LLMs and Symbolic Reasoning for C/C++ Vulnerability Analysis
NASA_VUL: A Comprehensive C/C++ Benchmark Dataset for LLM-Based Vulnerability Detection in NASA Software
P³R: Parallel Plugin-based Parameter Efficient Fine-tuning for Code Understanding through Hierarchical Representation Refinement
CSE 499 Senior Capstone Expo – Summer 2026
The CSE 499 Senior Capstone Expo showcased eight student group projects spanning cybersecurity, AI, and software engineering. Students designed and built real-world AI and software systems from the ground up — identifying meaningful problems, conducting original research, and delivering fully functional projects. The expo featured poster presentations, live demos, and 3-minute video pitches.
Expo Photos
Group Project Videos (3-Minute Pitches)
Group 1: PEFT – Vulnerability Detector
An AI-powered tool for identifying code vulnerabilities using parameter-efficient fine-tuning.
Group 2: Secure LPO
A framework for generating safer Python code through localized preference optimization.
Group 3: EventLens
An AI-assisted platform for capturing and sharing event moments across multiple platforms.
Group 4: Basketball AI Outcome
A full-stack application that predicts basketball game results using stacked ensemble machine learning.
Group 5: Multi-Agent RAG
A retrieval-augmented generation system powered by collaborative AI agents, with a focus on detecting stealthy poisoning attacks.
Group 6: DLN – Server Manager
A self-hosted Minecraft server management plugin built for a seamless user experience.
Group 7: DOIT
An intelligent gamified task and workflow automation assistant.
Group 8: Risk Reveal AI
An explainable phishing detection system combining deep learning, structural URL analysis, and LLM-based explanations.