I lead a research group working on machine learning for cybersecurity, and manage a portfolio of research projects for federal sponsors.
Machine learning has been central to my career. My early work focused on natural language processing, topic modeling, and large-scale knowledge graphs. Since 2008 I have applied machine learning to security, from feature selection and statistical modeling for production analytics to deep learning in PyTorch. My recent work focuses on large language models and agentic systems for cybersecurity.
Career at Livermore
- Applied machine learningNLP, topic modeling, security analytics
- Knowledge graphs and NLP systemsSoftware engineering and team lead
- Cybersecurity researchNetworks, malware, endpoints
- LLMs and agentic systemsApplied to cybersecurity
- Group LeaderLine management and team building
- Teaching and competitionsDOE CyberFire and CyberForce
- Principal InvestigatorMachine learning and cybersecurity programs
- 20032008201320182023Now
Education
University of Washington
MS, Computer Science
UC Davis
Doctoral coursework, Computer Science
Old Dominion University
BS with honors
Short bio
For event programs and speaker introductions. This text may be used without modification.
William Kallander leads a cybersecurity research group at Lawrence Livermore National Laboratory, where he directs applied machine learning research in cybersecurity for protecting critical infrastructure. Over more than two decades at the Laboratory he has worked across machine learning, natural language processing, knowledge graphs, malware, and network analysis, and his recent work focuses on large language models and LLM-based agents for cybersecurity. He is a principal investigator on multi-year machine learning and cybersecurity programs, teaches malware analysis for the Department of Energy’s CyberFire program, and designs challenges for the DOE CyberForce competition. He holds an MS in computer science from the University of Washington.
A full résumé is available on request.