Collaborating with Agents - Part 1

This three-part series is for learners who are already productive with AI and ready to take the next step. It moves beyond prompting and into directing: Part 1 introduces agents and reverse prompting, Part 2 covers agent skills and memory, and Part 3 focuses on designing multi-agent workflows that handle complex tasks. Participants leave with the confidence to direct AI agents and lead others through the same journey.

This first session begins with a background on generative AI and an introduction to AI agents. From there, participants use a secure sandbox to create their own autonomous agent and practice reverse prompting: guiding the agent to clarify goals, constraints, and missing information before acting. Participants leave with a working personal agent and a foundational understanding of how to direct it well.

• What LLMs are
• How copilots differ from agents
• Where agents fit in modern work
• How to guide an agent by asking it to clarify goals, constraints, risks, and missing content

• Explain the difference between chat AI and agent AI
• Set up and interact with a personal AI agent
• Connect an output to Zenbin and publish a web page

• Participants should have familiarity with at least one generative AI tool (e.g. ChatGPT, Gemini, Claude)
• This series builds on the skills introduced in the Introduction to AI and AI for Professionals classes.
• A free Telegram account is required. Install at web.telegram.org before class.