Concepts for AI Agentss
AI agent concepts are the ideas that separate an agent from a plain LLM call: the model picks its own next step inside a loop instead of following a fixed script. This first unit sets the vocabulary of workflows, agents, context windows and the tool-observe cycle that later units build on.
Stage: Foundations · Part of the AI Agents roadmap · Last reviewed 19 November 2026
What to learn in Concepts
- Explain the difference between a fixed workflow and an autonomous agent with one example of each
- Describe the loop of model call, tool call and observation
- Identify when a single prompt is enough and an agent is overkill
- Estimate how context window size and token cost limit a long agent run
- Name the failure modes of loops, such as repeated calls or drift
Concepts interview questions
What is the difference between a workflow and an agent?
In a workflow, your code defines the sequence of steps and the model fills in parts of it. In an agent, the model decides which step or tool comes next. Workflows are more predictable; agents handle open-ended tasks but cost more and fail in less obvious ways.
When should you not use an agent?
Avoid an agent when the steps are known in advance, latency or cost is tight, or errors are expensive. A deterministic workflow or one well-written prompt is cheaper, easier to test and easier to debug. Add autonomy only when the task path truly varies.
Where Concepts fits in the AI Agents roadmap
Concepts belongs to the Foundations stage (intermediate level), which takes about 23 hours. What an agent is.
Other units in this stage
Back to the full AI Agents roadmap