Putting Agentic AI to Work: From Everyday Tasks to Intelligent Automation

Agentic AI is more than automation with a language model attached. When AI can monitor progress, remember context, follow up, and choose its next step, it can help carry work forward without waiting for a person to prompt every action. This creates new possibilities for getting things done, as well as new responsibilities around judgment, oversight, and trust.

This hands-on workshop explores where agentic AI could make a practical difference in participants' own work and how to approach its use responsibly. The workshop moves from familiar everyday uses of AI toward systems that can act with greater independence.

Through guided exercises and examples drawn from participants' own roles, they will have the opportunity to:

- Identify tasks and goals that are good candidates for agentic AI.
- Distinguish between work that needs a fixed workflow and work that benefits from an agent able to choose its own path.
- Define the context, tools, permissions, and boundaries an agent needs.
- Decide how an agent's work should be reviewed and when a person must step in.
- Evaluate whether an agent is performing reliably enough to be given greater autonomy.
- Design a small, supervised agentic workflow for a real need in their work.
 

Concepts are paired with hands-on practice using currently available AI tools. Because these products are changing quickly, the workshop emphasizes durable methods rather than particular features. It introduces a framework for assessing new agentic tools, adapting them to one's work, and making informed decisions about where, and where not, to rely on them.

The workshop concludes with participants identifying a practical opportunity from their own work and sketching an initial design for an agentic solution, along with considerations for testing it safely and responsibly.