Foundations of Agentic AI: From Models That Respond to Systems That Act Online
AI is shifting from models that answer a question to systems that pursue a goal. Deciding what to do, taking action, observing the result, and adjusting. Think of an assistant that researches a question across many sources, or one that debugs code by running it and reading the errors. This tutorial lays the groundwork for understanding that shift: what makes a system "agentic," how the core reason–act–observe loop works with tools and memory, and the key patterns behind most agents. We'll also cover the parts that are easy to underestimate: when you actually need an agent versus a simpler workflow, how single agents extend to multi-agent systems, and what makes agents genuinely hard to build reliably, from evaluation to cost to safety. No framework experience needed, and this is the first in a series shaped by what you want to learn next.
Graduate students from Computer Science will lead this presentation.
This session will not be recorded.
Brought to you by the K-State AI Consulting Service (KAICS).
- Date:
- Monday, Oct 19 2026
- Time:
- 11:00am - 12:00pm
- Time Zone:
- Central Time - US & Canada (change)
- Campus:
- Virtual
- Online:
- This is an online event. Event URL will be sent via registration email.
- Audience:
- Faculty/Staff Students
- Categories:
- AI Literacy
