Agents in action

1. Agents in action

In this video you'll meet two built-in agents: Analyst and Researcher. Each one does a job the general chat does not.

2. Analyst Agent overview

The Analyst agent is built for deeper data work. It looks for patterns, trends and relationships in your numbers.

3. Analyst Agent overview

You'll find it in the Agent Store, under your agents. It works from natural language prompts, so you don't need formulas or complex queries.

4. Analyst Agent overview

For example, you can upload a spreadsheet of event registrations and ask: "Which region has shown the fastest growth compared to last quarter?" The key is that this agent works best with structured datasets like spreadsheets or CSV files, where information is organized into rows and columns. That structure lets Copilot analyze with precision. By using plain language, you stay focused on your business questions while the Analyst Agent does the heavy lifting.

5. Analyst Agent example

Let's use the EcoGreen_Registrations.xlsx file as our example. You might ask: "Find the top 3 regions by total registrations and show them in a chart. Walk me through your reasoning". The Analyst agent scans the file, groups the registrations by region, and builds the chart, all without you writing a single formula. You also see its reasoning steps laid out in the interface. That shows how it reached the answer, which makes the analysis more transparent and much easier to explain to your team. That last part matters more than it sounds. The slow half of analysis usually isn't getting the number, it's being able to defend it when someone asks where it came from. With the steps written out, you can answer that in the meeting instead of going back to rebuild the workbook afterwards.

6. Researcher Agent overview

The Researcher Agent is designed for in-depth information gathering. It does more than summarize. It reads across your documents and presentations and pulls out what you asked for. Standard summarizing condenses what is already there. The Researcher agent goes further and analyses it.

7. Researcher Agent overview

It identifies relationships between ideas and can highlight gaps or themes running across multiple files, which makes it useful for higher-level work such as campaign planning, background research or building strategic messaging. Its value lies in connecting material from several sources into a single line of reasoning you can work from.

8. Researcher Agent example

Researcher also does well with more open questions. You could ask: "Research the reusable water bottle market: 2026 trends, main competitors, and what buyers care about. Cite sources". Rather than answering straight away, it lays out a research plan and works through it step by step, gathering sources and weighing how well supported each finding is. Normally this would mean hours of searching, reading, and trying to connect the dots yourself. You'll get a structured, cited answer in minutes, so your time goes on strategy rather than searching.

9. Let's practice!

The Analyst Agent is your data expert, and the Researcher Agent is your information specialist. Together, they make working in Microsoft 365 faster and more accurate. Now it's time to practice what you have learned with some exercises.

Create Your Free Account

By continuing, you accept our Terms of Use, our Privacy Policy and that your data is stored in the USA.