Extending the Chatbot: Getting Started with AI Agents and Human-in-the-Loop Design

Curious how to put AI to work as a collaborative partner? AI agents are here – digital assistants that can plan travel, organize your schedule, or serve as a personal tutor.

What You’ll Learn

This two-part workshop introduces you to the world of AI agents and how humans have a required role within agentic AI in a practical, hands-on way. You will learn to…

• Define the core components of an AI agent.

• Recognize and explain the human role and function in making agents work.

• Create a functional AI Agent to perform a custom skill in a free LLM interface (such as Gemini or ChatGPT).

• Evaluate the effectiveness to identify improvements in logic or accuracy.

How You’ll Learn It

In the first session, you’ll move beyond simple chatting to explore the Agent Skills architecture. You will learn how to create an agent using the SKILL.md format—a lightweight way to give AI specialized knowledge and workflows. Together, we will learn how to build a "Skill Folder", complete with instructions and templates, and load these skills into free versions of ChatGPT, Claude, or Gemini.

Between sessions, you’ll practice building your own AI agent as a skill. You’ll have the opportunity to test it out, get feedback on it from the instructor, and iterate on it.

Finally, we’ll come back together to share the skills you’ve built, troubleshoot why certain instructions might "hallucinate," and discuss the pros and cons of agentic AI. By the end of this course, you will not just be using AI; you will be authoring your own digital teammates that can be improved, versioned, and shared. And most importantly, you’ll come to an understanding that without humanity, there is no agency.  

Who Should Attend

This course is ideal for curious professionals and adult learners who have dabbled with AI but want to level up their AI knowledge and skills and explore human-in-the-loop design. You don’t need to be a tech wizard, but you should know basic computer skills, including how to save and organize files.

What You'll Need

Participants need a laptop with a simple text editor installed (like Visual Studio Code, or Notepad or TextEdit) or a cloud-based editor (like Google Docs or VS Code for Web), as well as a free account with ChatGPT, Google Gemini, or Claude.

Workshop Details

When

Wednesdays, April 22, and May 6, 2026
6:15 - 7:45 PM
In between the two workshops, learners will work on building their own agent and have the opportunity to get asynchronous feedback on their work from the instructor.

Where

MSU Bozeman (Norm Asbjornson Hall, room 137)
Free parking is available after 6:00 pm in the Bobcat lots (#13 and 15 on this map).

Registration Fee

$100

A discount is available for nonprofit and public employees, as well as MSU staff, faculty and students. To obtain the code, email kelsye.gould@montana.edu with your affiliation.

Instructor

Jason A. Clark is a Professor and Head of Research Optimization, Analytics, and Data Services at Montana State University, where he has spent two decades working at the intersection of emerging technology, data science, and information systems. His research specializes in machine learning, explainable AI, reinforcement learning, and algorithmic UX design. He has led or co-led federally funded grants through IMLS, CLIR, and NLM, building AI tools for data discovery, responsible AI frameworks, and semantic knowledge systems. As Lead Software Developer on multiple grants, he brings rare practical depth alongside his scholarly credentials. Clark has presented on content-based filtering, knowledge graphs, and LLM-driven research optimization - all foundational to modern AI agents. As a longtime mentor and invited computer science lecturer, he excels at making complex agentic AI concepts accessible to practitioners across disciplines.

Questions

Contact Kelsye Gould, Continuing Education Program Manager, at 406-994-3062, kelsye.gould@montana.edu.