This is the part most people want to know about, so it is stated plainly. Your document is received by our API, forwarded to Microsoft Azure AI Document Intelligence for text extraction, and the extracted text is returned to you in the HTTP response.
We do not write the document, or the text extracted from it, to our database or to any persistent storage of ours. Once the response has been returned, our copy exists only in memory and is released.
What we do keep is the metadata listed above: the file name you supplied, its size in bytes, and the page count. We keep those because they are what your usage log and your bill are made of. We do not keep the contents.
We do not use your documents or extracted text to train any model, and we do not permit our processor to do so.