Collections¶

A collection is a group of files you can attach to a chatbot to give it access to content. There are two types of collections:

• Media collection - Send files — such as images, PDFs, video, or audio — to participants during a conversation.
• Indexed collection - Let your chatbot search your documents and ground its answers in that content (RAG).

Adding a collection to a chatbot¶

Navigate to the Collections section in the sidebar and click "Add new". Once the collection is created, you will be able to upload files to it.

After your collection has been created and populated with files, you can link it to any LLM node.

Collections and published chatbots¶

Collection content is a live shared resource: updates to your files in a collection reach your published chatbot automatically, without a republish. This applies whether you update a collection manually or via a scheduled document-source sync.

• Adding or removing files from a media collection takes effect for users immediately.
• Document-source syncs to an indexed collection — for example, nightly Confluence or GitHub syncs — are applied to the published chatbot as each sync completes.

The collection structure of a published chatbot version — which collections are linked to which pipeline nodes — is still frozen at publish time. To change which collections a chatbot uses, you must publish a new version.

For more detail on versioning in general, see Versioning.