About
Physicist turned machine learning engineer, now building apps and a company.
I started in physics at UC Irvine, studying energetic-ion loss in fusion plasmas, then moved through biomedical forensic science and computer science at Syracuse, where my research was on mission assurance and formal verification. Since then I have applied machine learning to computer vision at the Air Force Research Laboratory, signal demodulation at Geon Technologies, and graph neural networks for circuit analysis and hardware assurance at Chip Scan.
Sol of AI is my personal studio: the place I build and share small applications and take on freelance software work. I am also the co-founder and CEO of AI Security Assurance, which builds AI-driven tools to automate systems security analysis. For security analysis, visit AI Security Assurance and reach out through its Contact Us page.
Experience
Co-Founder & CEO
- Co-founded and lead an early-stage company building AI-driven tools to automate systems security analysis; principal investigator on DARPA and SBIR proposals. That work is done there, separately from Sol of AI.
Machine Learning Engineer
- Applied graph neural networks to circuit analysis for hardware assurance; evaluated over 40,000 models across heterogeneous prediction tasks over three years, managing 33–50 concurrent experiments with MLflow tracking.
- Built a multi-task model selection and serving application (Postgres, React); mentored interns and wrote monthly research reports for government stakeholders under SBIR Phase I and II contracts.
Machine Learning Engineer
- Investigated machine-learning approaches to signal demodulation, benchmarking Transformers, LSTMs, and BiRNNs with FFT-based preprocessing in Python and PyTorch on synthetic signal datasets.
Associate Computer Scientist
- Researched computer-vision methods (YOLO, Faster R-CNN, custom preprocessing) to extract mission-critical information from video text overlays, using a GAN to expand a small training dataset.
- Served as program manager for AI/ML research contracts and as technology liaison between AFRL and a joint forces organization; evaluated proposals, recruited research partners, and wrote funding proposals.
Graduate Research Assistant, Mission Assurance
- Researched STORM, a systems-theoretic operational risk management methodology, in collaboration with a US Army subject matter expert (M.S. thesis).
- Demonstrated Certified Security by Design with HOL theorem proving and secure state machines.
- Collaborated with Assured Information Security on a prototype that automates formal verification, building the compiler (in Haskell) that translates system concepts of operation into formally verifiable code.
Education
M.S. Computer Science
M.S. Biomedical Forensic Science
B.S. Physics and Applied Computational Mathematics
Publications
Context-Robust Identity Linkage in Social Media via Keystroke Dynamics