AI-900 Practice Question: Describe features of generative AI workloads on Azure

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✗

  A human employee who manages AI model deployments

  Why it's wrong here

  This is wrong because a human employee who manages AI model deployments is a person, not a software system. AI agents are autonomous software entities that operate without direct human intervention, using LLMs to plan and execute actions. While a human may oversee agents or perform MLOps, the term 'AI agent' in Azure AI Fundamentals specifically refers to an LLM-driven software component, not a role, job title, or human-in-the-loop operator.

• ✓

  An autonomous system using an LLM to plan and execute multi-step tasks using tools

  Why this is correct

  This is correct because an AI agent is a software system that combines an LLM's reasoning capacity with a feedback loop: it decomposes a user goal into a sequence of steps, selects appropriate tools (e.g., search engines, APIs, calculators, code interpreters), interprets tool outputs, and iterates until the goal is achieved. The LLM acts as the 'brain' that plans and adapts, while the tools extend the agent's ability to affect the real world or retrieve up-to-date information. This matches the AI-900 definition of an agent as an autonomous, LLM-driven task executor rather than a passive responder.

• ✗

  A monitoring agent that checks AI model health automatically

  Why it's wrong here

  This is wrong because health monitoring, such as tracking model latency, drift, or error rates, is operational telemetry—typically implemented with Application Insights, Azure Monitor, or custom schedulers. While a monitoring script might be automated, it does not use an LLM to reason about goals or plan multi-step tasks with tools; it simply checks predefined metrics and raises alerts. AI agents are characterized by goal-directed reasoning and tool use, not by status checks or threshold evaluation.

• ✗

  A software robot that scrapes websites for training data

  Why it's wrong here

  This is wrong because a web scraper is a data acquisition program that programmatically fetches and parses HTML, often for building training datasets—it performs no LLM-based reasoning, no task decomposition, and no dynamic tool selection. Scrapers follow fixed rules or selectors to extract content, whereas an AI agent evaluates context, formulates an action plan, and uses tools as means to accomplish an end. The absence of a reasoning loop with an LLM distinguishes simple extraction from genuine agentic behavior.

Go deeper

Related to this question