We weren't very happy with the state of AI based PDF/Image based document data extraction. Most solutions will digitize the header data as well as the company meta data and totals of an invoice for example but don't go the extra distance to make descriptive and line item level data actionable. They also charge for every new category of field and the accuracy isn't great. We needed a solution that automatically digitized every element of an invoice or trade document, classified the data, and made even the descriptive text actionable. With this update, we guarantee data accuracy at 100%.