.png)
🚀 Looking for a better way to integrate Flowcase with AI?
Check out the Flowcase Agentic MCP. It works with Copilot, Claude, Cursor, and any MCP-compatible tool. Easier to set up, more powerful, and actively maintained.
The guide below is still useful for legacy PowerApps connector setups.
Many of Flowcase's customers are leveraging their people and project data within Microsoft Copilot. This is enabling them to gather insights about their firm's expertise, and create proposal content faster than ever.
If you're also looking to use your Flowcase data within Copilot, we've created this guide that walks you through the complete setup process.
Prerequisites
Before starting, ensure you have:
- A Flowcase account
- A Flowcase API key
- Access to Microsoft PowerApps
- Access to Copilot Studio
Getting a Flowcase API Key
To obtain your API key, visit: https://docs.flowcase.com/authentication-security#overview--getting-an-api-key
Setting up the PowerApps Connector
Step 1:
Navigate to https://make.powerapps.com/ and click on the "...more" button to view all available data integration options.
Step 2:
Select "Discover all" from the menu options.
Step 3:
Under the Data card, locate and select "Custom connectors." Pin this option to your left-hand menu for easy access later.
Configuring the Connector
Click "New custom connector" in the top right corner and select "Create from blank."
Naming Your Connector
Give your connector a name. While "Flowcase" works, consider using something more descriptive if you plan to create multiple connectors.
Cleaning Data for Better LLM Results (Optional)
Note: If you skip this step, you may quickly exceed the context limit in the LLM-generated answers. However, if you plan to display results as cards instead of using the LLM to answer queries, this step might not be necessary.
Under the Code tab, you can add custom code to process the JSON output from Flowcase and filter out unnecessary data. For detailed information about the API response structure, visit: https://docs.flowcase.com/cvs-resume#overview--search-cv-résumé
Additionally, see this Microsoft link for more information: https://learn.microsoft.com/en-us/connectors/custom-connectors/write-code
Toggle the Swagger editor and paste the following code (replacing <subdomain> with your Flowcase subdomain):
Testing Your Connector
Click "New connection" and add your API key in the format: Bearer API_KEY. Add a query value in the freetext_search box
After creating the key, you'll be redirected to your connections. To test you have to navigate back to Custom connectors.
If your query returns JSON with CVs, your setup is successful!
Copilot Studio Integration
Setting Up Actions
Navigate to Copilot Studio and open either your agent or the main Copilot 365 instance. Go to the Actions page and click "Add an Action"
Search for the connector name you created earlier and select your connector from the results
If you completed the test step earlier, you should already have an authentication profile available by selecting the three dots on the right. If not, set it up in the format Bearer API_TOKEN
Decide whether authentication should be top-level or individual to Copilot users.
Set the "must" inputs as dynamic to allow the LLM to choose what to fill. The "size" parameter can be either hardcoded or set on a request basis
Click the "Add action" button to complete the setup
Test your action using the panel on the right side
Conclusion
This integration process can be customized to determine what data you want available in your agent. We're excited to see how you use our API with Copilot integrations!
For questions, feedback about this guide, or to showcase what you've built, contact us at techsupport@flowcase.com.
.png)
