Document - Data Source#

Use one or more documents as a data source. This section covers how to upload documents.

Click on Document to add documents as a data source:

Upload#

Upload your files (click or drag and drop on the dedicated area):

EmptyFiles selected

Extracting text from images and scans#

You can upload image files (.png, .jpg, .gif, .bmp, .webp, .tiff) and scanned PDFs as a data source, just like any other document. By default only the text that already exists in the file is indexed — text that lives inside an image is not read. To make that text searchable, turn on one of the two extraction modes below for the file before indexing.

Each row in the file list has a toggle on its right-hand side. OCR and LLM are mutually exclusive, so enabling one automatically turns the other off.

Optical Character Recognition (OCR)#

OCR converts a text image into machine-readable text. For example, scanning a document produces an image file that cannot be edited, searched, or word-counted directly.

OCR converts that image into a text document whose contents are stored as text data.

LLM parsing#

LLM parsing sends the image (or scanned page) to a vision-capable language model that reads it the way a person would. It generally handles complex layouts, tables, and handwriting better than OCR, and can also describe non-text visual content.

To enable a mode, click the OCR or LLM switch on the right-hand side of the file you want to apply it to:

Once all documents are specified, click "Finish". The data source page opens:

Batch Actions on Selected Files#

You can select multiple files at once using the checkboxes on the left side of each row.

Once one or more files are selected, a batch action toolbar appears above the table with the following actions:

• Download — downloads all selected PDF files at once. The button is active only when at least one selected file is a PDF or audio file in Indexed status.
• Delete — permanently removes all selected files from the connector.