Nominations for the 2022 KMWorld Readers’ Choice Awards Extended to July 5

Intelligent data capture: a trend only beginning Part 1

Brainware has a strong presence in the oil and gas marketplace with customers like BP, Shell, Conoco and Halliburton. It attained that presence by leveraging its relationships with large integrators including BearingPoint and CapGemini.

Brainware's A/P-distiller product is designed to handle large invoice volumes and particularly shines in table extraction. The Brainware Table Extraction Engine is an intelligent engine that extracts tables and associated line item information from invoices using line pattern recognition technology. The line item detail can be validated or used to collect additional data elements through table-lookups. A/P-distiller excels by using advanced learning technology to "read" and "interpret" documents for sorting and data extraction. It can process multiple and changing invoice types; extract critical data and line item detail; and export data to existing financial, workflow and ECM systems--significantly reducing the costs associated with document sorting, manual data entry and validation.

Brainware's Order-to-Cash distiller (OTC-distiller) virtually eliminates manual data entry of sales orders and requires no changes to established business practices. It automatically sorts incoming documents, extracts all required data from a purchase order and creates new orders in the backend order processing system. OTC-distiller handles multipage orders, multilingual documents and an unlimited variety of purchase order formats.

Document Strategies is a research-oriented firm that uses advanced pattern recognition, neural and AI technologies to create a simple-to-use, yet powerful capture and search technology set. It has unique capabilities and strengths when recognizing unstructured documents.

President Jim Bowen says, "Our customers can set up rules for unknown document types and pull out key information--without any pre-defined templates. Our AI engine 'understands' the context of content in the documents, therefore finding information and patterns buried deep in thousands of pages of documents." Its largest group of customers are in the legal vertical market segment in discovery/litigation support applications.

Document Strategies has a marked strength in recognizing poor-quality documents--those with as low as 50 percent recognition confidence rating--such as water-damaged or very old documents, and faxes that may be fourth generation. Its capture technology maintains formatting, and search technology allows users to retrieve search requests quickly and to modulate the accuracy level (and therefore resultant number of hits) of searches. That is especially helpful in litigation support applications. Users may refine searches even using similar "thesaurus" terms or synonyms (e.g. look for "corporation" and also find incidences of "business"). Due to its unique technology, it performs very fast searches, typically sub-second, which is unusual for AI technology in general.

EMC Captiva has traditionally focused on larger deals in the higher end of the marketplace. Today, the focus is moving downstream to more middle-market companies, but being able to process unstructured and semi-structured documents using IDC brings about a wealth of new opportunities.

"Capture technology has evolved to a point that semi-structured documents can be processed like structured forms used to be," says Rob Jensen, EMC Captiva's director of corporate communications. "That means we can go after new applications in large accounts." The neural technology acquired from absorbing SWT allows Captiva software to "learn" an invoice format so that the next time an invoice from that vendor is processed, it is instantly recognized.

Beyond Documentum, EMC Captiva has partnerships with major ECM providers like IBM, FileNet, Open Text and Hyland. Its largest integration partner is EDS (eds.com), and it sells approximately 40 percent of its deals through the indirect channel. Captiva's IDC offerings are aimed at the lower page volume range and are based on SWT's b-Wize product line. They include b-Wize MAIL to capture and classify incoming paper and electronic documents from one location; b-Wize MONEY to capture checks, payment slips, deposit forms, etc.; b-Wize INVOICE and b-Wize FORM. Additional products include InputAccel for Invoices and Digital Mailroom.

DICOM's Kofax unit has traditionally focused on the lower end of the marketplace and has built a vast network of channel partners—more than 1,100 worldwide including healthcare giant Cerner, as well as SYSCOM, FileNet, Hummingbird and IBM Global Services.

"Our market has traditionally been post-archival, but now we are getting pulled forward in the business process, capturing data to feed business systems," says Sameer Samat, Kofax's VP of engineering and technology. "It's opened up a whole new set of opportunities for us."

Kofax INDICIUS is a suite of IDC modules that provide advanced document classification, separation and extraction capabilities, enabling the automated processing of more types of business documents. Some of its target applications include application form processing, mortgage processing, records management and mailroom.

KMWorld Covers
for qualified subscribers
Subscribe Now Current Issue Past Issues