Copilot Studio vs Microsoft Copilot
By Ricardo Costa Val do Rosario helped by Microsoft Copilot 365 and ChatGPT 5,0 Plus
1. Definition
1. Microsoft Copilot: AI assistant integrated into the Microsoft 365 ecosystem, which
uses organizational data and language models to increase productivity (e.g., generating
texts, analytics, summaries).
2. Copilot Studio: Platform for creating, customizing, and managing copilots, allowing
integration with internal systems, workflows, and company-specific data.
2. Structure
# Microsoft Copilot:
1. Based on AI models (GPT).
2. Integrated with Microsoft Graph and apps (Word, Excel, Teams).
3. It works as a productivity layer.
# Copilot Studio:
1. Low-code/no-code development interface.
2. Connectors to APIs and internal data.
3. Tools for customization and governance.
3. Functions
# Microsoft Copilot:
1. Automatic redaction (Word, Outlook).
2. Data analysis (Excel).
3. Meeting summary (Teams).
# Copilot Studio:
1. Creation of specific copilots for internal processes.
2. Complex tasks automation.
3. Integration with corporate systems.
4. Applications with Real Cases
# Microsoft Copilot:
1. Generation of financial reports in Excel.
2. Summary of emails and meetings for managers.
# Copilot Studio:
1. Creation of a copilot for internal technical support.
2. Automation of HR processes (e.g., onboarding).
5. Code Examples
- Copilot Studio allows scripts and connectors. Example in Power Automate:
# Example API call to integrate Copilot Studio
const response = await fetch("https://api.copilotstudio.com/query", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ prompt: "Generate sales report" })
});
const data = await response.json();
console.log(data.result);
6. Advantages and Disadvantages
# Microsoft Copilot
1. Easy integration with Microsoft 365.
2. Increases productivity without complex setup.
3. Limited to standard application logic.
4. Less customization.
# Copilot Studio
1. High customization and integration with internal systems.
2. Allows you to create process-specific copilots.
3. It requires basic technical knowledge.
4. Increased initial effort for setup.
7. Codes samples for Copilot Studio and Microsoft Copilot:
1. Example for Copilot Studio
- Copilot Studio allows you to create custom copilots using connectors and APIs. A
simple JavaScript example to integrate with the Copilot Studio API:
# Example API call to integrate Copilot Studio
const response = await fetch("https://api.copilotstudio.com/query", {
method: "POST",
headers: { "Content-Type": "application/json" },
body: JSON.stringify({ prompt: "Generate sales report" })
});
const data = await response.json();
console.log(data.result);
# What does it do?
1. Sends a prompt to the copilot via API.
2. Receives the processed response (for example, a sales report).
3. Can be integrated into a Power Automate flow or internal app.
2. Example for Microsoft Copilot (Excel)
Using Office Scripts to automate tasks with Copilot:
function main(workbook: ExcelScript.Workbook) {
let sheet = workbook.getActiveWorksheet();
let salesData = sheet.getRange("A1:B10").getValues();
Simulation: Send data to Copilot for analysis
console.log("Analyzing Sales Data:", salesData);
Here you could integrate with an endpoint that calls Copilot
}
What does it do?
1. Captures data from a spreadsheet.
2. Prepare to send to Copilot to generate insights or graphs.
3. Example with Power Automate
Creating a flow that sends data to Copilot Studio and returns a response:
{
"type": "Http",
"method": "POST",
"uri": "https://api.copilotstudio.com/query",
"headers": {
"Content-Type": "application/json"
},
"body": {
"prompt": "Summary of the week's meetings"
}
}
7. Applications of Copilot Studio in Healthcare
# Contextualization
- Copilot Studio, when contextualized for the health area, is no longer just
an automation tool and starts to act as a clinical-operational copilot, capable
of integrating electronic medical records, hospital systems, care protocols, and
administrative flows, respecting the requirements of LGPD, information security,
and technovigilance. Below, I present three real application contexts, with respective
examples of flows and codes,
