OpenBox

OpenBox features AI technology that utilizes OCR and ICR to automatically classify documents, extract data, and create dynamic indexes based on preset business rules. The system has a high tolerance for image quality issues, such as skew, distortion, scaling, dithering, density problems, and artifacts, resulting in higher accuracy and recognition, even with subpar images. If the confidence level is below a configurable threshold, the workflow will route that document to a manual data capture queue.

Using a variety of classification engines, OpenBox offers a comprehensive approach to data capture. These engines include:

• Self-Learning Classification - OpenBox uses training reference images created for each document class, allowing it to classify documents by comparing the original form image with reference images.

• Fuzzy Matching - This technique allows the system to match and correct data extraction based on probabilistic thresholds of approximate string matches.

• Print Type Classification - By detecting higher lines and more variety between lines, OpenBox is capable of capturing handwritten text.

• Statistical Classification - Free-form classification can learn and recognize the characteristics of a document type based on examples.

• Geometric Classification - The system can identify and classify documents based on simple or complex forms, including barcodes, text, numbers, logos, blocks, and lines.

Current intelligent capture solutions require document pre-sorting and separator sheets between multiple-page documents, contributing to long preparation times that account for up to 40% of the total input time.

OpenBox’s patented document separation technology applies a unique stamp on the first page of multiple-page documents to separate them automatically. This eliminates the need for separator sheets, and supports up to 7 levels of document structure with multiple methods of structure detection. By eliminating presorting and document separators, OpenBox helps you reduce labor, materials, and IT costs.

Exela’s technology automates data mining, classification, aggregation, and summarization from structured and unstructured sources. AI powers analytics and modeling engines that extract insights from large data sets. 

Automated queries, keyword triggers, defined alerts, and manual search parameters enable data capture from paper documents, websites, social media, news media, open communications, and internal sources. With all this information easily available, you’re better able to discover patterns, identify correlating factors, and utilize predictive modeling to anticipate future trends.