Michael Vertin
Software Engineer
I am a software engineer with a foundation in computer science and mathematics, focused on building efficient and reliable systems. I enjoy applying mathematical concepts such as probability, algorithms, and data structures to design solutions that are theoretically sound and practically effective. In several projects and competitions, I leverage concepts like probability, number theory, and time complexity to create efficient solutions using analytical reasoning to go beyond standard approaches.
Skills & Technologies
Featured Projects
SCA Image Search (Cline Library)
AI-powered image search tool developed to help researchers explore 100,000+ archival documents in Northern Arizona University's Special Collections and Archives.
- Exhibited leadership and teamwork roles to implement feature extraction and inner product similarity for fast, accurate image retrieval.
- Deployed containerized backend and frontend services via Docker Compose and AWS EC2.
- Delivered a production-ready application supporting large-scale academic research.
Robot Invasion
Real-time tower defense game featuring dynamic wave difficulty scaling.
- Designed turrets with unique abilities to challenge players with adaptive gameplay.
- Balanced game mechanics to require strategic planning and timing.
- Demonstrates object-oriented game design and real-time system management.
High-Performance Word Search
Created a word search solver that outperformed all student and instructor submissions in a class-wide competition.
- Achieved 500% faster performance on the largest test board and 1300% faster on the smallest.
- Implemented traversal pruning, probabilistic skipping, and configuration-specific optimizations.
- Ranked #1 in the class-wide competition by significant margin.
Competitive Halma AI
Built an automated Halma game-playing agent that won a bracket-style class tournament.
- Implemented dynamic depth adjustment (5–6) with randomized depth extension to meet time limits.
- Enhanced alpha-beta pruning with shallow-sorting heuristics for faster move evaluation.
- Outperformed all opponents, most of whom were limited to static depth-4 searches.