Discord
The Discord Source Connector allows you to integrate Discord channels and messages as a data source for your pipelines. This guide explains the configuration options available when setting up a Discord source connector.
Before you begin
Before starting, you'll need:
The Server ID for your Discord server.
Your Discord bot's token.
The Channel ID for the Discord channel you want to connect to your RAG pipeline.
Configure the connector
To configure a connector to your Discord instance:
Click Source Connectors from the main menu.
Click New Source Connector from the Source Connectors page.
Select the Discord card.
Enter connection parameters in the form using the Discord Parameters table below as a guide, then click Create Discord Integration.
When you specify your configured Discord connector by name in your pipeline configuration, Vectorize ingests all messages up to 10,000 messages (by default) in the specified channel that match the parameters you've set.
Discord Parameters
Field | Description | Required |
---|---|---|
Name | An alphanumeric name of your choosing. Used to identify this connector in your pipeline configuration. It does not need to match any other parameter in configuration. | Yes |
Server Id | The Discord server Id. | Yes |
Token | The Discord bot's token. | Yes |
Channel ID | Your Discord channel ID. - To add additional channels, click Add and add each channel ID in the new text box which appears. | Yes |
Discord pipeline configuration
Adjust the Discord pipeline configuration parameters if needed, using the Discord Pipeline Configuration Parameters table below as a guide, or leave them at the default settings.
Click Save Configuration when done.
Discord Pipeline Configuration Parameters
Field | Default value | Notes |
---|---|---|
Emoji Filter | none | This filter restricts data ingested by your pipeline to messages where someone responded with this specific emoji. This filter only works with custom emoji you've added to your Discord server; it does not work with standard emoji. - Enter the name of the emoji, not the emoji itself; for example, planet (not :planet:). - To add additional emoji, click Add and add the name of each emoji in the new text box which appears. |
Author Filter | none | This filter restricts data ingested by your pipeline to messages sent by one or more specific users. - Enter a username (not a display name). - To add additional usernames, click Add and add each username in the new text box which appears. |
Ignore Author Filter | none | This filter restricts data ingested by your pipeline to messages not sent by one or more specific users. - Enter a username (not a display name). - To add additional usernames, click Add and add each username in the new text box which appears. |
Limit | 10000 | The number of messages that will be processed per channel during your pipeline's initial backfilling stage. There are no limits on how many messages will be processed after the pipeline has backfilled. |
Thread Message Inclusion | ALL | Controls how all messages in a thread are read. - ALL: every response in a thread will be consumed as long as the message that starts the thread passes all filters. - FILTER: each response in a thread must also pass all the filters. |
Filter Logic | AND | Controls the logic applied to all filters. Options: - AND - OR Example: "has the planet emoji and is authored by user123" vs. "has the planet emoji or is authored by user123" |
Thread Message Mode | CONCATENATE | Determines how message threads are consumed by Vectorize. - CONCATENATE: All the messages in a thread are treated as one message entity, where each response in a thread is concatenated into one message, but retains its conversation format. (For this mode, only the message that starts the thread has to pass the filters.) - SINGLE: Every response in a thread is treated as its own message. |
Troubleshooting
Confirm browser access. Confirming that you have access to the specified channel through your browser.
Test connectivity outside of Vectorize. Try confirming accessibility to your account from your own client or API tool, like Postman.
Try a new token. Regenerate your token from within the Discord UI, and update its value in your connector. You can also try connecting to a different channel to further isolate the issue.
What's next?
If you haven't yet built a connector to your vector database, go to Configuring Vector Database Connectors and select the platform you prefer to use for storing output vectors.
OR
If you're ready to start producing vector embeddings from your input data, head to Pipeline Basics. Select your new connector as the data source to use it in your pipeline.
Last updated