Getting the most from your uploaded documents.
While Zudello can extract most document types there are a few things to keep in mind to ensure a faster and more accurate extraction.
While this section will go into more detail the key areas to look out for are:
-
Low-quality photos
-
Handwritten documents
-
Data covered by stamps & writing
-
Smudged paper documents scanned.
Supported File Types
While Zudello prefers PDF files it also supports a range of other document types including Images (JPG, JPEG and PNG) and CSV/Excel/Word documents.
Scanned Documents and Photos
Uploading an image of your document is a really easy way to transfer from your desk into Zudello. In order to extract the correct fields, it is important the image is as clear as possible.
Here are a couple of quick tips to get the perfect extraction.
-
Take the photo as straight as possible.
-
Use the focus on your camera to ensure the information isn't blurry.
-
Take a photo of the entire receipt/document.
-
Avoid writing over the data on the document.
Zudello will extract as much information as possible placing it into the corresponding fields. Any unread data will then trigger a User review for accuracy.
Scanning is also a great way to upload your documents to Zudello. Much like images, the result of extraction from scanned documents is based on the quality of the document.
Problems are often caused by:
-
Smudges on the document.
-
A severely skewed scan.
-
Writing or other marking on the document.
-
Highlighted documents (with a black & white scan).
Handwritten documents
Zudello’s ability to extract information is maximised when handwriting is both neat and legible. A good rule of thumb here is if you can't clearly make out the words Zudello won't be able to either.
If you need to process handwritten documents BLOCK LETTERS are much clearer than smaller cursive letters.
CSV and Excel
Invoices and other documents received in CSV or Excel format can be processed by Zudello. Column headings can be trained in, however, Zudello will automatically detect the column headings of each document. The closer they match field names in Zudello the quicker they will extract.