n8n YouTube Transcription and De-duplication Workflow

This n8n workflow automates the process of fetching new YouTube videos from a specified channel, transcribing them using an external API, and storing the transcriptions in a Supabase database. It includes de-duplication logic to avoid processing the same video multiple times.

What it does

This workflow streamlines the transcription of YouTube content by:

1. Scheduling a Check: Periodically (e.g., daily, hourly) checks a YouTube channel's RSS feed for new video uploads.
2. Extracting Video IDs: Parses the RSS feed to get the YouTube video IDs of recently uploaded videos.
3. De-duplicating Videos: Queries a Supabase database to check if a video with the extracted ID has already been transcribed.
4. Filtering New Videos: Proceeds only with videos that have not yet been processed.
5. Requesting Transcription: For each new video, it sends a request to a transcription API (e.g., Transcript.io) to get the video's transcript.
6. Formatting Data: Prepares the video metadata and transcription for storage.
7. Storing in Supabase: Inserts the video ID and its transcription into a Supabase table.
8. Error Handling: Includes a "Stop and Error" node, suggesting potential error handling or notification for failed transcription attempts or database operations.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance.
• YouTube Channel RSS Feed URL: The RSS feed URL for the YouTube channel you want to monitor.
• Transcription API Key: An API key for a transcription service (e.g., Transcript.io, as suggested by the directory name, though the JSON uses a generic HTTP Request node).
• Supabase Account: Access to a Supabase project with a database table configured to store video IDs and their transcriptions.
  • You will need your Supabase Project URL and Anon Key.
  • The Supabase table should have at least two columns: one for the YouTube video ID (e.g., video_id) and another for the transcription (e.g., transcript).

Setup/Usage

1. Import the Workflow:
   • Download the provided JSON file.
   • In your n8n instance, go to "Workflows" and click "New".
   • Click the "Import from JSON" button and paste the workflow JSON or upload the file.
2. Configure Credentials:
   • HTTP Request (Transcription API):
     • Locate the "HTTP Request" node (ID: 19) that calls the transcription API.
     • Configure the URL, method (likely POST), and any necessary headers (e.g., Authorization with your API key) or body parameters required by your chosen transcription service.
     • You might need to dynamicall pass the YouTube video URL or ID to this API.
   • Supabase:
     • Locate the "Supabase" node (ID: 545).
     • Click on "Credentials" and add a new Supabase API credential.
     • Enter your Supabase Project URL and Anon Key.
3. Configure Nodes:
   • Schedule Trigger:
     • Adjust the "Schedule Trigger" node (ID: 839) to your desired interval for checking new videos (e.g., every hour, daily).
   • RSS Read:
     • In the "RSS Read" node (ID: 37), set the "URL" field to the RSS feed URL of the YouTube channel you want to monitor.
   • Edit Fields (Set):
     • Review the "Edit Fields" nodes (e.g., ID: 38) to ensure the data is being transformed as expected for your Supabase table and transcription API.
   • Supabase (Read):
     • In the "Supabase" node that checks for existing videos, ensure the "Table" and "Filters" are correctly configured to query your transcription table using the video_id.
   • Supabase (Write):
     • In the "Supabase" node that inserts new transcriptions, ensure the "Table" and "Column values" are correctly mapped to your Supabase table's schema.
   • Code Node:
     • The "Code" node (ID: 834) likely contains custom logic for extracting video IDs or preparing data. Review and adjust it if your RSS feed structure or data requirements differ.
4. Activate the Workflow:
   • Once all configurations are complete, save and activate the workflow. It will now run automatically based on your schedule.