Skip to content
TariahAcademy
← Catalog

Agentic AI · Operator

Agentic AI Foundations

How AI agents plan, call tools and recover from errors — and when not to use one.

Lessons

  • 1. What makes a system agentic

    Agents vs. chatbots vs. workflows; autonomy as a dial.

    25m

  • 2. Tools and function calling

    Describing tools so a model uses them correctly.

    40m

  • 3. Memory, state and context windows

    Short-term scratchpads vs. long-term stores.

    35m

  • 4. Failure modes and human-in-the-loop

    Loops, hallucinated tools, and approval gates.

    30m

Exercises

Answers save in this browser. Compare yours with the checklist — no automated grading yet.

Write a tool spec

Define a JSON schema for a `book_meeting` tool with name, attendees and time.

Spot the failure

An agent retries a failing API 40 times. List two safeguards.