Use the Power Automate Webhook Integration to let your Noem.AI chatbot trigger Microsoft Power Automate flows from natural-language instructions. The bot calls an HTTP trigger in your flow with a JSON payload you define and can surface the flow’s response to the user.
What this action does #
- Sends an HTTP request (POST/GET) from the bot to a Power Automate flow that starts with an HTTP trigger.
- Passes Data Properties (JSON) gathered from the conversation.
- Optionally displays values returned by the flow (confirmation number, URLs, etc.) in chat.
Prerequisites #
- A Power Automate Cloud flow that begins with “When an HTTP request is received” (Instant trigger).
- The flow is Saved so the HTTP POST URL and Request JSON schema are available.
- Any required Microsoft 365/Dataverse/SharePoint connectors in the flow are authorized.
Create the Power Automate action #
- Open: Actions → Add → Power Automate to open Power Automate Webhook Integration.
- Name (required)
The function name you’ll reference in the bot’s Instructions.
Example: pa_submit_support_ticket. - Description (required)
Explain when the bot should call this action and which fields to collect.
Example:
“Use pa_submit_support_ticket when a user reports an issue. Gather title, detailed description, priority, product area, and user email. Confirm before sending.” - End Point(required)
- Method: POST (recommended) or GET.
- URL: Paste your flow’s HTTP POST URL from the trigger.
- (Optional) Hide Function Log in Chat to suppress request logs in the transcript.
- Method: POST (recommended) or GET.
- Data Properties
Define the fields the bot includes in the request body (POST) or query string (GET). For each property set:
- Name (e.g., emailContent, priority)
- Type (String, Number, Boolean, Array, Object)
- Description (what the AI must capture)
- Required (must be present before calling)
- Name (e.g., emailContent, priority)
- Use Add New Property to add fields. AI Assistance can suggest properties from your Description.
- Advanced Mode (optional)
Configure custom headers (e.g., Authorization or a shared secret), query parameters, and response mapping if your flow expects them. - Save
Click Save to store the action.
Tip (payload inspection): While designing, you can temporarily point the End Point to https://webhook.site/ to see the exact JSON your bot will send, then switch to your flow URL.
Designing the Power Automate flow #
- Trigger: Add When an HTTP request is received.
- Define the request schema: Paste a JSON sample matching your Data Properties and let Power Automate generate the schema.
- Process the data: Add actions (e.g., create a SharePoint item, Dataverse row, send Teams message, create Planner task).
Respond to the bot (optional): Add Response action with a JSON body such as:
{ “status”: “created”, “ticketId”: “SUP-4821”, “portalUrl”: “https://support.example.com/tickets/SUP-4821” }
Example configuration #
- Name: pa_submit_support_ticket
- Description:
“Call when users report bugs or request help. Collect: short title, detailed description with steps, priority (low/medium/high), product area, reporter email. Confirm before sending.” - End Point: POST https://prod-XX.westus.logic.azure.com:443/workflows/…
- Data Properties:
| Name | Type | Description | Required |
| title | String | Short issue title (≤100 chars). | ✓ |
| emailContent | String | Detailed description including steps, expected vs. actual, and context. | ✓ |
| priority | String | One of: low, medium, high. | |
| productArea | String | Feature/module where the issue occurred. | |
| reporterEmail | String | User’s email (use {{user.email}} if present; otherwise ask). | |
| attachments | Array | URLs to screenshots/files provided by the user. |
Sample JSON (POST body):
{
“title”: “Checkout 500 error”,
“emailContent”: “Steps: add item > Pay -> 500. Expected: success.”,
“priority”: “high”,
“productArea”: “Checkout”,
“reporterEmail”: “[email protected]”,
“attachments”: [“https://files.noem.ai/f/abc123.png”]