Building AI Agents for ICS/OT Security: From Zero to Orchestrator
About This Session
(Full day (8 hours with breaks - $495 Fee)
AI agents are transforming how we approach critical infrastructure security, but most teams are still using basic chatbots, not even realizing the full potential of what AI agents can really do. In this hands-on workshop, you’ll build real AI agents that can reason about ICS/OT and cybersecurity, remember context across engagements, coordinate multi-step workflows, and more. You’ll learn to do it safely and securely.
Starting from scratch, participants will:
• Set up and agent environment
• Build agents that understand ICS/OT protocols, threat landscapes, regulatory frameworks, and more
• Create persistent memory systems so agents learn from every engagement instead of starting fresh
• Test their agents against realistic scenarios
• Learn how to use AI agents to be more efficient with your time, save time on tasks, improve time to delivery, and supercharge just about every task you do
No prior AI engineering or programming experience required, just a laptop and curiosity. You’ll leave with a working agent framework you can customize for your own environment.
What Makes This Different
We’re not teaching prompt engineering. We’re teaching agent engineering — how to build AI systems that think, remember, and coordinate like a seasoned security team. Using Claude Code sub-agents as the foundation, you’ll learn the safest, most practical path to deploying AI in critical infrastructure environments.
Prerequisites
Laptop with terminal access, GitHub account, Anthropic API key (free trial available). Basic command-line familiarity helpful but not required.
AI agents are transforming how we approach critical infrastructure security, but most teams are still using basic chatbots, not even realizing the full potential of what AI agents can really do. In this hands-on workshop, you’ll build real AI agents that can reason about ICS/OT and cybersecurity, remember context across engagements, coordinate multi-step workflows, and more. You’ll learn to do it safely and securely.
Starting from scratch, participants will:
• Set up and agent environment
• Build agents that understand ICS/OT protocols, threat landscapes, regulatory frameworks, and more
• Create persistent memory systems so agents learn from every engagement instead of starting fresh
• Test their agents against realistic scenarios
• Learn how to use AI agents to be more efficient with your time, save time on tasks, improve time to delivery, and supercharge just about every task you do
No prior AI engineering or programming experience required, just a laptop and curiosity. You’ll leave with a working agent framework you can customize for your own environment.
What Makes This Different
We’re not teaching prompt engineering. We’re teaching agent engineering — how to build AI systems that think, remember, and coordinate like a seasoned security team. Using Claude Code sub-agents as the foundation, you’ll learn the safest, most practical path to deploying AI in critical infrastructure environments.
Prerequisites
Laptop with terminal access, GitHub account, Anthropic API key (free trial available). Basic command-line familiarity helpful but not required.
Speaker
Clint Bodungen
Director, AI/ML Engineering - Arcova
Clint Bodungen is an OT cybersecurity pioneer and AI engineer with 30+ years of experience. Known for pioneering practical, innovative approaches to cybersecurity, he has spent the last two decade at the forefront of integrating gamification and AI into cybersecurity and training. He has published two books, Hacking Exposed: Industrial Control Systems and ChatGPT for Cybersecurity Cookbook, and created ThreatGEN® Red vs. Blue, an online cybersecurity simulation game, as well as AutoTableTop, an AI-driven incident response platform that uses AI to build and run IR tabletop exercises. Today, as Director of AI/ML Engineering at Arcova, his work centers on creating production-ready AI systems that help transform and modernize the cybersecurity industry.