Work with AI agents

  • 5 minutes

So far, you've seen how AI can help answer questions, summarize information, generate content, and analyze data.

In many cases, a single request and response are enough to help you complete a task. However, some work requires multiple steps, multiple sources of information, or coordination across several activities.

This is where AI agents can help.

What is an AI agent?

An AI agent is an AI-powered assistant that can help complete a task or achieve a goal by working through a series of actions.

You can think of an agent as an AI assistant that helps move work forward toward a specific goal.

Tip

While a chat interaction often focuses on a single request and response, an agent can help coordinate multiple activities as part of a larger process.

AI chat versus AI agents

Many people begin using AI through chat-based experiences such as asking for a meeting summary or a presentation outline. These are examples of direct interactions where the AI responds to a specific request. Agents build on these capabilities by helping manage a broader task or workflow.

For example, an agent might:

  • Gather information from multiple sources.
  • Analyze the information.
  • Create a draft proposal.
  • Identify risks or open questions.
  • Recommend next steps.

Instead of focusing on a single response, the agent helps support an entire process.

Diagram of AI agents surrounded by surface insights, organize information, and perform routine tasks.

How agents can help at work

Screenshot of office workers looking at a laptop.

Organizations often have recurring activities that involve multiple steps and multiple people.

Examples include:

  • Preparing customer proposals
  • Reviewing project status
  • Researching a market opportunity
  • Onboarding new employees
  • Managing customer support requests
  • Tracking project milestones

Agents can help organize information, perform routine tasks, and surface insights that help people make decisions.

The goal isn't to replace people. Instead, agents help reduce manual effort and allow people to focus on the work that requires judgment, expertise, and collaboration.

AI agents still require human oversight

Although agents can help automate parts of a process, people remain responsible for the outcome.

Before acting on information provided by an agent, it's important to:

  • Review the results
  • Verify important information
  • Check recommendations
  • Apply business knowledge and judgment

AI can assist with work, but people remain accountable for decisions and actions.

Tip

Think of an AI agent as a teammate that helps gather information and complete tasks. Like any teammate, the work still benefits from review, guidance, and feedback.

Scenario: Use an agent for research

Diagram of researcher agent.

In the previous unit, you learned how grounding helps AI generate responses based on relevant work information.

Now Jo needs to prepare materials for an executive review of the Contoso website redesign project. Rather than collecting information and creating materials manually, she asks the Researcher agent to help coordinate the work.

Define the goal 🎯

Jo asks the Researcher agent:

Prepare a draft presentation for an executive review of the Contoso website redesign project. Summarize project status, identify risks, and recommend next steps.

Screenshot of the Researcher agent prompt and referenced files.

Review the results 🔍

The Researcher agent gathers information from project documents, customer communications, meeting notes, and other relevant sources. Jo answers a few follow-up questions, and Researcher creates a draft presentation that includes:

  • Project status
  • Key risks
  • Open questions
  • Recommended next steps

Note

The Researcher agent can take time to gather and analyze information, so be patient.

Screenshot of the briefing Researcher agent drafted.

Prepare for review 🗂️

Jo now has a draft presentation she can review, refine, and adapt for the executive audience.

Instead of manually gathering and organizing information across multiple tasks, the Researcher agent helped Jo create a strong starting point more quickly.

Screenshot of the presentation draft created by the Researcher agent.

Tip

Agents can help coordinate multiple activities as part of a larger workflow, but people remain responsible for reviewing results and making decisions.

Why it matters

AI agents extend the capabilities of AI by helping coordinate tasks that involve multiple steps. Rather than responding to a single request, an agent can gather information, analyze findings, create content, and support progress toward a specific goal.

As demonstrated in Jo's proposal scenario, agents can help reduce manual effort and support complex workflows. However, people remain responsible for reviewing outputs, applying judgment, and making decisions. Agents help move work forward, but human oversight remains an important part of the process.

Next, let's look at how to use AI responsibly and with confidence.

Note

We recognize that different people like to learn in different ways. You can choose to complete this module in video-based format or you can read the content as text and images. The text contains greater detail than the videos, so in some cases you might want to refer to it as supplemental material to the video presentation.