Register now! 

Timing

This course comprises two (2) live Zoom sessions.

Session 1: July 15 from 1 - 4 pm ET 

Session 2: July 16 from 1 - 3 pm ET

Unable to make the live sessions? Recordings will be made available (automatically to those who register) within a week after the last session takes place, though continuing education credits can only be provided for attending live events.

Event Overview

AI is moving into utility workflows whether organizations plan for it or not. This two-session course gives cybersecurity practitioners a working understanding of how modern AI systems operate, where they create risk, and how to bring AI into a utility environment under the organization’s own controls. Participants learn to evaluate AI tools, write governance that holds up in practice, and build a private, self-hosted AI assistant trained on their own data. That assistant runs entirely in-house and never sends sensitive information to an outside service.

The course pairs concepts with hands-on practice. By the end, participants will have worked through governance frameworks, the AI-enabled attack techniques now being aimed at utilities, and a full build: preparing training data, fine-tuning a model, deploying it locally, and tuning its behavior.

Who Should Attend

This course is built for cybersecurity staff at public power utilities and electric municipalities: security analysts, IT and OT administrators, and the managers who will own AI policy. It assumes general IT and security familiarity. It does not assume any prior machine learning experience. Anyone who has been asked “what is our position on AI?” and needs a concrete answer will get the most from it.

What Participants Receive

  • A practical governance framework they can adapt for their own organization, including acceptable use, a prohibited-data list, and approved-tool criteria.

  • A working knowledge of current AI-enabled attacks against utilities and the controls that reduce that exposure.

  • A complete, repeatable process for building and running a private AI assistant on their own hardware.

  • Instructor-provided reference material, command references, and the training notebook used in the hands-on build.

Prerequisites

  • General IT and cybersecurity familiarity, such as networking, endpoints, and basic security controls.

  • No prior AI or machine learning experience is required.

  • Comfort with a command line helps for the hands-on build but is not required to follow along.

Materials and Setup

Participants will need:

  • A laptop or workstation for the hands-on portions.

  • A free Kaggle account, which is used for model training.

  • The ability to run a virtual machine, or a spare machine for the Linux host. The build uses Ubuntu, Docker, and Ollama.

  • To run a small model locally, roughly 16 to 24 GB of RAM and a GPU with 8 GB of VRAM (a 12 GB NVIDIA card is preferred). The course also covers lighter-weight options.

Hands-on participation is encouraged but optional. Participants can follow the build during the sessions and finish it afterward using the provided materials.

Speaker

Travis Cleek is a Partner at SkyHelm and the Chief Technology Officer at GridVantage. He has roughly 18 years of experience across IT, cybersecurity, and operational technology.

His career began as an IT Director at a multinational manufacturing firm. He then moved to an Oklahoma electric cooperative, where he strengthened the network and security infrastructure with a focus on SCADA communications and smart grid work. While at the cooperative, he co-founded a fiber-to-the-home ISP and led its early business development and network design. He also worked alongside electrical engineers to help develop the world’s first real-time self-healing grid system.

Travis’s work centers on cybersecurity and operational technology, particularly SCADA systems, and on making complex infrastructure easier to secure and operate. Over the past two years he built a private AI tool for use by a 24/7 utility-focused security operations center. That work led to GridVantage, a company he launched with a utility, which offers a privately trained AI tool that handles complex electrical engineering and operations tasks.

At SkyHelm, Travis helps electric cooperatives and public power utilities secure their networks and operational technology environments. His background runs through systems and network administration, robotics, and cybersecurity, with earlier roles in higher education, manufacturing, and the utility sector.

Technology Requirements

  • Zoom Platform: This event will be hosted on Zoom. A reliable internet connection is essential for smooth participation. You can test your setup at Zoom.us/test .
  • Audio: We recommend using your computer’s microphone and speakers for the best experience. A dial‑in option will also be available.

Login Credentials and Recording

  • Zoom Links: Access details will be emailed 24 hours before the event, with a reminder sent one hour prior to the start time.
  • Recording: The event will be recorded and distributed to registrants after the meeting.