How to Draft email responses with Power Automate

My flows > New flow > Instant cloud flow

Create new instant cloud flow

Log into make.powerautomate.com and click My flows in the left sidebar. Click New flow and select Instant cloud flow. Name it 'Email Response Drafter' and choose 'For a selected message' as the trigger. This creates a Slack bot command that team members can trigger on any message.

1. 1Click 'My flows' in the left navigation
2. 2Click the 'New flow' button
3. 3Select 'Instant cloud flow' from the options
4. 4Name the flow 'Email Response Drafter'
5. 5Choose 'For a selected message' trigger

Trigger > Search connectors > Slack

Add Slack message trigger

Click the trigger block and search for 'Slack' in the connector list. Select 'Get message' action and authenticate your Slack workspace. You'll need to grant permissions to read messages and channels. The trigger will capture the message content when someone right-clicks and selects your flow.

1. 1Click the trigger block to expand options
2. 2Search for 'Slack' in the connector search bar
3. 3Select 'Get message' from Slack actions
4. 4Click 'Sign in' to authenticate your workspace
5. 5Grant message and channel read permissions

New step > Compose > Data Operations

Parse message content

Add a 'Compose' action to extract the email content from the Slack message. Click New step and search for 'Compose'. In the Inputs field, select 'Text' from the Slack trigger dynamic content. This captures whatever email text was pasted into Slack for processing.

1. 1Click 'New step' below the Slack trigger
2. 2Search for 'Compose' in the action list
3. 3Select 'Compose' from Data Operations
4. 4Click in the Inputs field
5. 5Select 'Text' from Slack dynamic content

New step > HTTP > Premium

Add OpenAI connection

Click New step and search for 'HTTP' since Power Automate doesn't have a native OpenAI connector. Select the HTTP action and set the method to POST. You'll configure this to call OpenAI's chat completions endpoint directly using your API key.

1. 1Click 'New step' after the Compose action
2. 2Search for 'HTTP' in the connector list
3. 3Select 'HTTP' from Premium connectors
4. 4Set Method to 'POST'
5. 5Enter URI: https://api.openai.com/v1/chat/completions

HTTP action > Headers > Body

Configure OpenAI API call

In the HTTP action, add headers for Content-Type and Authorization. Set Content-Type to 'application/json' and Authorization to 'Bearer YOUR_API_KEY'. Replace YOUR_API_KEY with your actual OpenAI API key. In the Body field, you'll structure the JSON request with the system prompt and user message.

1. 1Click 'Add new parameter' and select 'Headers'
2. 2Add header: Content-Type = application/json
3. 3Add header: Authorization = Bearer sk-your-api-key-here
4. 4Click in the Body field
5. 5Switch to code view and paste the JSON structure

HTTP action > Body

Structure the AI prompt

In the Body field, create a JSON structure that tells OpenAI to draft a professional email response. Include a system message defining the assistant's role and a user message containing the original email from Slack. The system prompt should specify tone, length, and response format requirements.

1. 1Click in the Body field of the HTTP action
2. 2Select 'Expression' tab in dynamic content
3. 3Build JSON with system and user message roles
4. 4Reference the Compose output for email content
5. 5Set model to 'gpt-3.5-turbo' for cost efficiency

New step > Parse JSON > Data Operations

Parse OpenAI response

Add a 'Parse JSON' action to extract the response text from OpenAI's API response. Click New step and search for 'Parse JSON'. In the Content field, select 'body' from the HTTP action dynamic content. You'll need to provide a sample JSON schema that matches OpenAI's response format.

1. 1Click 'New step' after the HTTP action
2. 2Search for 'Parse JSON'
3. 3Select 'Parse JSON' from Data Operations
4. 4Set Content to HTTP response 'body'
5. 5Click 'Generate from sample' for schema

New step > Slack > Post message

Post response to Slack

Add another Slack action to post the drafted response back to the same channel. Search for Slack connectors and select 'Post message'. Set the channel to the original message's channel using dynamic content. Include both the original email and the AI-generated response for easy copying.

1. 1Click 'New step' after Parse JSON
2. 2Search for 'Slack' and select 'Post message'
3. 3Set Channel to the original message channel
4. 4Format the message with original email and response
5. 5Add thread_ts to reply to the original message

Flow designer > Save > Test

Test the workflow

Save the flow and test it by going to your Slack channel and pasting a sample client email. Right-click the message and look for your flow name in the context menu. Select it to trigger the automation. The response should appear as a threaded reply within 10-15 seconds.

1. 1Click 'Save' in the top right of the flow designer
2. 2Go to your Slack workspace
3. 3Paste a test email in a channel
4. 4Right-click the message and select your flow
5. 5Wait for the AI response to appear

Add this JSON structure to the HTTP Body field to create more contextual responses that reference the sender's name and specific details from the original email.

{
  "model": "gpt-3.5-turbo",
  "messages": [

{
  "model": "gpt-3.5-turbo",
  "messages": [
    {
      "role": "system",
      "content": "You are a professional assistant drafting email responses. Be polite, concise, and reference specific details from the original message. Keep responses under 150 words."
    },
    {
      "role": "user",
      "content": concat('Please draft a professional response to this email: ', outputs('Compose'))
    }
  ],
  "max_tokens": 200,
  "temperature": 0.7
}