Microsoft Copilot Studio

Connect Akto with Microsoft Copilot Studio

Overview

Microsoft Copilot Studio is a platform for building and deploying conversational AI agents and copilots. This setup is recommended if you want to monitor API traffic from your Copilot Studio applications and ensure your AI-powered conversations maintain security standards.

The Akto Copilot Studio connector automatically:

• Fetches all Copilot Studio conversation data

• Monitors bot interactions and user conversations

• Sends API traffic data to Akto for security analysis

Steps to Connect

1

Configure Akto Traffic Processor

Set up and configure Akto Traffic Processor. The steps are mentioned here.

2

Set Up Application Insights

To monitor your Copilot Studio application, you need to configure Azure Application Insights:

1. Create Application Insights Resource (if not already created):

   • Follow the Azure Application Insights documentation to create a new Application Insights resource

2. Configure Your Copilot in Copilot Studio:

   • Follow the Copilot Studio Application Insights integration guide to connect your copilot to Application Insights

3. Get Your Application Insights Credentials:

   • Follow the API Access documentation to get your Application ID and create an API Key with "Read telemetry" permission

3

Clone the Akto Infrastructure Repository

Clone the Akto infrastructure repository:

wget https://raw.githubusercontent.com/akto-api-security/infra/refs/heads/feature/quick-setup/docker-compose-copilot-studio-cron.yaml

wget https://raw.githubusercontent.com/akto-api-security/infra/refs/heads/feature/quick-setup/copilot-studio-cron.env

wget https://raw.githubusercontent.com/akto-api-security/infra/refs/heads/feature/quick-setup/watchtower.env

4

Configure Copilot Studio Environment Variables

Update the following variables in the copilot-cron.env file:

APPINSIGHTS_APP_ID=<YOUR_APPLICATION_INSIGHTS_APP_ID>
APPINSIGHTS_API_KEY=<YOUR_APPLICATION_INSIGHTS_API_KEY>
AKTO_KAFKA_BROKER_URL=kafka1:19092

5

Start the Copilot Studio Traffic Connector

Run the following command to start the Copilot Studio traffic connector:

docker compose -f docker-compose-copilot-cron.yaml up

This will start monitoring your Copilot Studio conversations and send API traffic data to Akto for analysis.

What Data is Collected?

The Copilot Studio connector collects conversation data from your copilot applications:

Conversation Data

• All bot conversations and user interactions

• User messages and bot responses

• Conversation topics and flows

Metadata

• Conversation IDs and timestamps

• Channel information (Web Chat, Teams, etc.)

• Activity IDs for tracking message pairs

How It Works

The connector uses Azure Application Insights to retrieve conversation telemetry:

1. Data Source: Application Insights stores all conversation events from your Copilot Studio bots using the Bot Framework

2. Polling: The connector polls Application Insights every 10 minutes (configurable) to fetch new conversation events

3. Processing: Events are paired (user message → bot response) to create complete conversation flows

4. Publishing: Processed data is sent to Akto's Kafka topic for security analysis

Supported Channels

The connector works with all channels supported by Copilot Studio:

• Direct Line (Web Chat)

• Microsoft Teams

• Copilot Studio Test Chat

• Custom channels

All conversation sources are normalized by the Bot Framework before being logged to Application Insights.

Get Support for your Akto setup

There are multiple ways to request support from Akto. We are 24X7 available on the following:

1. In-app intercom support. Message us with your query on intercom in Akto dashboard and someone will reply.

2. Join our discord channel for community support.

3. Contact [email protected] for email support.

4. Contact us here.