A Futuristic View of the Auto-Coding Industry


A Futuristic View of the Auto-Coding Industry

  • 1 Tags

By Sean Keefer


The time? Circa 1950. The problem? A large number of documents need to be organized, indexed, and reviewed for the purposes of discovery and ultimate production. Issues such as privilege, waiver, redaction, and even basic identification and scope must be confronted. The solution? Dive in headfirst and be thankful that half a century ago the norm in the majority of cases was that you were dealing with fewer rather than more documents.

Now move the clock forward. Just about two decades ago, if a litigation support professional was confronted with the above task, the options were not much different than those that existed in 1950. By the early 1990’s, there was more paper to be dealt with and the same issues of privilege, waiver, redaction, and document identification still had to be addressed. Today, at the turn of the 21st century, the picture has changed even more. The norm for litigation involves dealing not only with mountains of paper, but electronically stored information (ESI) as well. The task can best be characterized as Herculean.

The idea of coding documents electronically instead of manually was so cutting edge in the early 1990’s that many litigation support professionals were reluctant to accept it. It may even have been thought of as “alien.” Early on, those litigation support professionals that did choose the “new” approach usually had to work with a third-party vendor to get the documents into electronic form. Then an attorney, paralegal, or the third party vendor would review each document image, determine its value, and code key information into a database. This procedure was time consuming and costly, but the results that it produced have forever changed the way documents are now being handled in litigation.

Back then, cases involving a massive number of documents required a great deal of space for storage and precious time to index and physically locate individual documents. Now documents can be made available in both electronic and searchable forms. In short, the process is becoming more automated. The use of human coders, with the advent of auto-coding software, can now be reduced but not entirely eliminated. With the advent of document management databases, one can store tens of thousands, or even millions of documents and get almost instant search and access capability. Given the time and cost savings, as well as the ever-growing number of options available by having the documents in electronic form, it is no surprise that the process of auto-coding has quickly caught on.

Technology and a New Direction

The first iterations of auto-coding delivered to service bureaus that would then consult with the litigation support professional. Coding parameters would be determined, the documents would be processed for OCR, and then coded databases would be delivered back after some period of delay. The expense was usually great and the end product often provided little more than basic data.

However, the industry is undergoing a shift. Auto-coding technology that was previously available only through service bureaus can now be licensed for efficient and economical use to individual firms, corporations, governmental agencies, and other businesses–in desktop form. This has created even more in the way of tools to make the job of the litigation support professional more efficient.

This shift to desktop form has also created a change in the way that auto-coding technology is marketed. Previously if a law firm, corporation, governmental entity, or any other prospective user wanted to take advantage of auto-coding technology, it was off to the service bureaus where a long wait ensued and a large bill followed. However, the approach taken by auto-coding vendors now is to deliver the technology to the users to allow them to process and code their documents in-house, thereby cutting out the service bureau middlemen. “Auto-coding technology was really cost prohibitive to many firms. The result was that many people who could have benefited from the technology couldn’t afford to use it,” said Lisa Rosen of Rosen Technology Resources, Inc., a re-seller of auto-coding software.

According to Joe Howie, Director of Product Marketing for Syngence, LLC, “The present focus of the auto-coding industry is to take the software from the service bureaus and put it in the hands of the end-users.” This approach gives the end-user more control and also makes for a quicker turnaround in document processing. Something else that end-users will appreciate about this shift is that it helps keep the ultimate price down.

“Auto-coding technology was really cost
prohibitive to many firms…” said Lisa
Rosen of Rosen Technology Resources, Inc.,
a reseller of auto-coding software.”

By using this approach, according to Howie, the end-user can process the documents as they need them and when they need them. Certain documents can be targeted for processing, documents can be processed in small or large batches, all documents can be processed, or in any custom manner the user may decide. The end result is that the enduser has the control over the approach to the document processing. It allows them the freedom to manage the case support in a more efficient fashion doing as much or as little as need be.

Great Impact on Productivity and Efficiencies & Compliance

According to Lisa Rosen, however, “There’s confusion with many people in regard to auto-coding and what it can provide. Some people feel that auto-coding will only provide the metadata from the documents, but this is not the case. Metadata doesn’t provide the total amount of information about the document.” Metadata is simply basic information about the data – how large the file may be, the date it was created or modified, and other related information about the document, not the information contained within the document. Auto-coding provides more than metadata; it provides data specific to the individual documents, such as beginning/ending bates number, page count, estimated and/or document date, title string, document type, author, addressee, names mentioned, organizations, dates mentioned, keywords mentioned, batch name, other number, along with a document “confidence score” that allows users to gauge the “quality” level of the resulting auto-coding.

An example of document processing where the use of auto-coding can have a great impact is when dealing with issues such as privilege and waiver. Under the new Federal Rules of Civil Procedure, privilege and waiver can be important concepts of which litigation support professionals need to be keenly aware. It is quite possible that when large numbers of documents are produced, a privileged document could inadvertently be included. While the new Federal Rules contemplate this issue and provide a mechanism to potentially retain the privilege, the difficulty is avoided entirely if the document is never produced in the first place. By using components available within some auto-coding products, it is possible to take measures to limit the exposure in this situation.

“The philosophy is that the end client can be
responsible for the use and customization
of the product… ” said Howie.

Some auto-coding software products not only identify the master document, but also mark drafts, e-mails, working copies, copies in memorandums, and any other instance of the document. These applications make it possible to identify the universe of privileged documents prior to production to better ensure that privilege will not be waived. This process also allows for simultaneous management of both electronic documents and paper documents that have been converted to electronic format via OCR. As this example illustrates, the latest auto-coding products focus on end-user efficiency and quantification and control becoming the guiding principle. “The philosophy is that the end client can be responsible for the use and customization of the product. From page level to end analysis, the more the client can take control of it, the better the product will work for them,” said Howie.

The Future is Now

As the future marches forward, the auto-coding industry is focusing on the integration of other applications into the processing pipeline to provide as many options to the litigation support professional as possible. As the technology becomes more available to the end users and the traditional service bureau becomes less the norm, the focus becomes giving the user more control by allowing them to integrate the databases created through the coding process into other software programs. “As auto-coding technology becomes more widely distributed, the approach to the software will become one of the best practices for litigation support,” according to Howie.

Keith Rowand, who has an extensive background in legal technology and is the architect of ALCoder™, indicates that while the learning curve on auto-coding products now seems to be great, in the future auto-coding will become but a boring part of the process and an integral link in the chain of efficient and streamlined litigation support.

When word processing and database programs first became available, they became standard for litigation support. When a firm hired new litigation support professionals, the products were standard and new professionals didn’t have to be trained in their use. In the best practices model, auto-coding, and auto-coding integration will enter this realm. Litigation support professionals will know how to use the standard programs, which will help to create the best practices scenario. “Auto-coding software is another tool in the litigation support arsenal,” said Rowand. By having data coded into database form, the possibilities for use are endless, well beyond organization and ease of retrieval. As auto-coding software companies work to improve the user interface of their products and also increase the options for integration into additional programs, the end-users will reap the benefits in both possibilities and cost savings. The auto-coding process is so full of potential. Keith Rowand summed it all up very well when he said, “Essentially, by using auto-coding to its full potential, users will have a tool box instead of a hammer to process data.”

About the Author

Sean Keefer is a freelance writer and attorney currently located in Charleston, South Carolina. Prior to shifting his professional focus to writing, Sean practiced as a litigator with a focus on complex domestic relations and alternative dispute resolution.

Reprinted with permission from Litigation Support Today magazine, May/July 2007. Copyright 2007 by Litigation Support Today. Originally published in the May/July 2007 issue of Litigation Support Today. Reprinted with permission for this edition only of The Reporter. Not intended for reproduction or public distribution. For other reprint permissions of any kind (electronic or print) contact Litigation Support Today at www.litigationsupporttoday.com