AI Voice/Text Note-Taking with Line Messaging, Supabase Vector DB & Gmail

This n8n workflow streamlines the process of capturing voice or text notes, processing them with AI, storing them in a Supabase vector database, and optionally sending them to Gmail. It's designed to make note-taking intelligent, searchable, and integrated with your communication and storage tools.

What it does

This workflow automates the following steps:

1. Receives Notes via Webhook: It listens for incoming voice or text notes through a webhook, likely from a messaging platform like Line.
2. Transcribes Voice Notes (if applicable): If an audio file is received, it uses OpenAI's Whisper to transcribe the audio into text.
3. Processes Text with AI: The transcribed text (or direct text input) is then processed by an OpenAI Chat Model to extract key information or summarize the note.
4. Filters for Actionable Content: An 'If' node checks if the AI-processed note contains a specific keyword (e.g., "email") to determine if it should be sent to Gmail.
5. Stores Notes in Supabase: The processed note, along with its embedding (generated by OpenAI Embeddings), is stored in a Supabase vector database for efficient semantic search.
6. Sends Notes to Gmail (Conditional): If the note is flagged for email, it's sent to a specified Gmail address.
7. Manages Chat History (Optional): It uses a Postgres Chat Memory to maintain conversational context, which can be useful for follow-up interactions.

Prerequisites/Requirements

To use this workflow, you will need:

• n8n Instance: A running n8n instance.
• OpenAI API Key: For transcription (Whisper), AI processing (Chat Model), and embeddings.
• Supabase Account: A Supabase project with a configured vector database. You'll need your Supabase URL and API Key.
• PostgreSQL Database (for Chat Memory): A PostgreSQL database for storing chat history if you intend to use the Postgres Chat Memory node.
• Gmail Account: Configured as an n8n credential if you want to send notes to Gmail.
• Webhook Source: A service (e.g., Line Messaging, Telegram, custom application) configured to send data to the n8n Webhook URL.

Setup/Usage

1. Import the Workflow: Import the provided JSON into your n8n instance.
2. Configure Credentials:
   • Set up your OpenAI API Key credential for the OpenAI, OpenAI Chat Model, and Embeddings OpenAI nodes.
   • Configure your Supabase credential for the Supabase Vector Store node, providing your Supabase URL and API Key.
   • If using the Postgres Chat Memory, set up your PostgreSQL credential.
   • If sending emails, configure your Gmail credential.
3. Configure Webhook:
   • Activate the Webhook node and copy its unique URL.
   • Configure your external service (e.g., Line Messaging webhook settings) to send POST requests to this URL when a new note (text or audio) is created.
4. Customize AI Agent (Optional): Review and adjust the prompts or tools within the AI Agent node if you want to modify how the AI processes your notes.
5. Define Email Keyword: In the If node, adjust the condition if you want to use a different keyword to trigger sending notes to Gmail.
6. Activate the Workflow: Once all credentials and configurations are set, activate the workflow.

Now, when you send a note (voice or text) to your configured webhook, the workflow will automatically process, store, and potentially email it.