To upload a file to the knowledge base:
1 Open the Editor for the relevant Blueprint or experience. For example, as shown below:
2 Select the Settings tab as shown above. The following is displayed:
2 Select the AI Assistant option in the left navigation panel, as shown below:
The AI Assistant pane is displayed, as shown below:
3 In the Knowledge Base section, add the files that the AI Assistant should use when answering participant questions. To learn more, see CloudShare AI Privacy and Security.
4 In the AI Assistant pane, click the Upload Files button, as shown below:
2 Select a file that is accessible from your computer and click the Open button.
The following file processing statuses apply to each file uploaded to the knowledge base:
- UPLOADING/UPLOADED: The file has been uploaded but has not yet been indexed by the AI Assistant, so it is not yet available to the AI Assistant.
- PROCESSING: The file is being indexed and is not yet available to the AI Assistant.
- READY: The file has been indexed successfully and will be made available to the AI Assistant after the Save button is clicked, as described below.
- FAILED: The file could not be processed. Remove the file and re-upload to retry.
Note: If there is a problem with the file format (for now, only PDF is supported), a relevant message is displayed.
3 After the files are READY (as shown above), click the Save button at the top right to upload the file.
Note: Before saving your changes, make sure the updates are correct and ready to be applied to the live experience. If the experience is live, the saved changes are applied immediately and may affect participant environments.
Note: File processing is triggered by the Save action, not by the upload itself.
Note: The AI Assistant will not use the file until processing is complete.
Note: The list of previously uploaded files remains visible.
4 Click the Save button at the top right.
Note: It is recommended to enable the Enable AI Assistant switch only after the relevant files have been uploaded and have the READY status. This ensures that participants can only access the AI Assistant after the knowledge base is ready to provide accurate, useful answers.