Add Documents

Overview#

Add Documents uploads local files into the Knowledge Base so the research agent can learn from your source material. Drag-and-drop or standard file selection both feed the same queue for processing.

When to Use It#

Importing policies, manuals, or spreadsheets you want the AI to reference
Bulk uploading multiple documents during initial workspace setup
Refreshing an existing folder with the latest version of a document

Key Benefits#

Bulk-friendly — drop up to ten files at once instead of uploading individually
Format aware — supports PDFs and spreadsheets commonly used in operations
Progress feedback — visual indicators show file-by-file upload and overall completion

UI/UX Specification#

Component Reference#

Component	Type	Description
Drag-and-drop zone	Upload area	Displays `Drag & Drop Files Here, Or Click To Select Files` instructions
Helper text	Text	Lists supported formats (`PDF, XLS, XLSX`) and the 100 MB maximum size
File queue	List	Shows each file name, type, size, and upload status/progress
Remove icon	Button	Deletes a selected file from the queue before processing
Queue button	Primary button	Confirms completed uploads and returns control to the knowledge hub

How to Use#

In the Knowledge Base tab, click + New and choose Add Source, then select the Document option.
Drag files into the upload zone or click the zone to open your file picker and select up to ten documents.
Wait for each file to finish uploading. Completed items show a green outline and 100% progress.
Remove any unwanted file using the trash icon before queuing.
Press Queue to confirm. The modal closes and the new sources appear in the knowledge base list for training.

Notes & Tips#

Each file must be 100 MB or smaller; larger files trigger an error and stay in the queue with a red state.
PDFs should contain selectable text. Scanned images without OCR cannot be ingested.
Excel files upload as text content only; formulas, charts, and styling are ignored during training.
If you add the wrong document, remove it before pressing Queue—there is no undo after submission.

FAQ#

Which file types are supported?#

The uploader accepts PDF, XLS, and XLSX files as indicated in the interface. Additional formats are planned for future releases.

Can I monitor upload progress?#

Yes. Each file card displays an individual percentage, and the header shows how many files succeeded.

What if a file fails to upload?#

Failed items stay in the queue with a red border. Fix the file (for example, reduce size) and re-upload before pressing Queue.

Do uploads start training automatically?#

No. After queuing sources, return to the knowledge hub to trigger training or assign the content to an agent.

Last updated on October 7, 2025