How to Translate Slack Messages with Pipedream

Pipedream > New > Build a workflow

Create new workflow

Navigate to pipedream.com and click the green 'New' button in the top right. Select 'Build a workflow' from the dropdown menu. You'll land on the workflow builder with a trigger step already created. Click on the trigger step to configure it.

1. 1Click 'New' in the top right corner
2. 2Select 'Build a workflow' from the dropdown
3. 3Click on the default trigger step
4. 4Search for 'Slack' in the app selector

Trigger > Slack > New Reaction Added

Configure reaction trigger

Select 'New Reaction Added' from the Slack trigger list. This event fires whenever someone adds an emoji reaction to any message in your Slack workspace. You'll need to connect your Slack account by clicking 'Connect Account' and going through Slack's OAuth flow.

1. 1Select 'New Reaction Added' from the event list
2. 2Click 'Connect Account' button
3. 3Authorize Pipedream in the Slack popup
4. 4Select your Slack workspace from the dropdown

Trigger > Generate Test Event

Test the trigger

Click 'Generate Test Event' to create sample data for your workflow. If no test data appears, go to any Slack channel and add a flag emoji reaction to a message. Return to Pipedream and click 'Generate Test Event' again. The test data shows the structure of reaction events.

1. 1Click 'Generate Test Event' button
2. 2If empty, go react to a Slack message with any emoji
3. 3Return to Pipedream and click 'Generate Test Event' again
4. 4Verify the test data includes reaction and message details

Add Step > Filter

Add filter step

Click the + button below the trigger to add a new step. Select 'Filter' from the step types. This filters reactions to only flag emojis that represent languages. Configure the filter to check if the reaction emoji matches flag patterns like 'flag-us', 'flag-jp', 'flag-de'.

1. 1Click the + button below the trigger step
2. 2Select 'Filter' from the step types
3. 3Set condition to 'Reaction contains flag-'
4. 4Add the expression: {{steps.trigger.event.reaction}}.includes('flag-')

Add Step > Code > Node.js

Add emoji mapping step

Add a 'Node.js Code' step to map flag emojis to language codes. This step converts Slack's flag emoji names like 'flag-us' or 'flag-jp' into language codes that OpenAI understands like 'English' or 'Japanese'. The mapping ensures accurate translation requests.

1. 1Click + to add another step
2. 2Select 'Node.js Code' from the options
3. 3Name the step 'Map Flag to Language'
4. 4Paste the language mapping code in the code editor

Add Step > OpenAI > Chat

Add OpenAI step

Click + to add an OpenAI step. Select 'Chat' action to use GPT for translation. Connect your OpenAI account using your API key from platform.openai.com. Configure the system message to instruct GPT to translate text and the user message to include the original Slack message text.

1. 1Click + to add a new step
2. 2Search for and select 'OpenAI'
3. 3Choose the 'Chat' action
4. 4Connect your OpenAI account with API key
5. 5Set model to 'gpt-3.5-turbo' for cost efficiency

OpenAI Step > Messages Configuration

Configure translation prompt

In the OpenAI step, set up the system and user messages. The system message should instruct GPT to translate text accurately while preserving formatting. The user message combines the target language from step 5 with the original message text from the trigger. Include context about keeping @mentions and #channels unchanged.

1. 1Set System Message: 'Translate the following text accurately. Preserve @mentions and #channels.'
2. 2Set User Message: 'Translate to {{steps.map_flag_to_language.language}}: {{steps.trigger.event.item.message.text}}'
3. 3Set temperature to 0.1 for consistent translations
4. 4Set max tokens to 500

Add Step > Slack > Send Message to Channel

Add Slack reply step

Add another Slack step to post the translation as a thread reply. Select 'Send Message to Channel' action. Configure it to reply in thread by setting the thread timestamp to the original message's timestamp. This keeps translations organized under the original message.

1. 1Click + to add a Slack step
2. 2Select 'Send Message to Channel'
3. 3Set channel to {{steps.trigger.event.item.channel}}
4. 4Set thread_ts to {{steps.trigger.event.item.ts}}
5. 5Set text to include the translation and source language

Deploy > Test in Slack

Test the complete workflow

Click 'Deploy' to activate your workflow. Go to a Slack channel where the Pipedream bot has access and post a test message. React to that message with a flag emoji like 🇫🇷 or 🇪🇸. The workflow should trigger within 2-3 seconds and post a translation as a thread reply.

1. 1Click the green 'Deploy' button in the top right
2. 2Go to a Slack channel with the bot invited
3. 3Post a test message in English
4. 4React with a flag emoji like 🇫🇷
5. 5Check for the threaded translation reply

This Node.js code maps Slack flag emoji names to full language names for better OpenAI translation accuracy. Paste it in the emoji mapping step between the filter and OpenAI steps.

export default defineComponent({
  async run({ steps, $ }) {
    const flagToLanguage = {

export default defineComponent({
  async run({ steps, $ }) {
    const flagToLanguage = {
      'flag-us': 'English',
      'flag-gb': 'English',
      'flag-es': 'Spanish',
      'flag-fr': 'French',
      'flag-de': 'German',
      'flag-jp': 'Japanese',
      'flag-cn': 'Chinese',
      'flag-kr': 'Korean',
      'flag-it': 'Italian',
      'flag-pt': 'Portuguese',
      'flag-ru': 'Russian',
      'flag-nl': 'Dutch',
      'flag-ar': 'Arabic',
      'flag-in': 'Hindi',
      'flag-br': 'Portuguese'
    };
    
    const reaction = steps.trigger.event.reaction;
    const targetLanguage = flagToLanguage[reaction] || 'English';
    
    // Add context for better translations
    const isBusinessContext = steps.trigger.event.item.message.text.includes(
      'meeting', 'project', 'deadline', 'report', 'budget'
    );
    
    return {
      language: targetLanguage,
      context: isBusinessContext ? 'business' : 'casual',
      original_emoji: reaction
    };
  }
});