Having control of your feature settings allows you to decrease your document processing time. Each feature we offer requires a specific amount of time to execute, so removing features you don't want to use (text-to-speech, detecting watermarks, etc..) can speed up your results!

You can find your flow feature settings under options; edit this flow.

Under the feature settings section, you also have the option to restrict document types for performance. This is helpful if you want every document read by a specific model type.

For instance, if you need every document read as an invoice, you will select invoice under document type, and each response will be read as such. This excludes our AI's ability to classify it as something else, such as a receipt or bill.

Feature Settings Definitions

The Document Object Model (DOM) is a specific data type in OCR that consists of elements such as pages, blocks, lines, words, and their corresponding location information.
The text-to-speech reader offers audio output for OCR data.
Detect faces, barcodes, signatures, watermarks, checkboxes, ink signatures, digital signatures, glare, and blur. Disabling any of these detection features will reduce processing time, but you will not receive an output deselected option in the responses. So, for instance, if you are parsing digital documents, you won't need glare or blur detection, or if there are no faces or watermarks, those can be disabled as well.
For verify document shape, we crosscheck the document for the correct classification. For example, a driver's license has to have a rectangular card-like shape.
Enabling segmentation will allow you to processes multiple documents within a single page
Recognize entities will specifically extract names, locations, events, and products from your documents when used in combination with semantic processing
Semantic processing will extract data from documents we do not have a pre-exisiting model for.
Render line items as objects convert line items into an array of objects
Flattening tables to fields allows you to convert a table in a document into individual fields, represented by the row and column headers

Customer Onboarding Guide

Getting Started