Deloitte has one of the largest Forensic practices in the world with over 1400 dedicated practitioners in over 30 countries. Their services are increasingly sought by companies and lawyers who require detailed investigations to be undertaken or disputes to be resolved. The Forensic group includes forensic accountants, legal and law enforcement specialists, and business intelligence experts. Their teams use state-of-the-art forensic technology to ensure that data is handled with maximum efficiency.

Due to the nature of eDiscovery work, forensic analysts must work with many kinds of documents and file types. Their job is to comb through this data and select the appropriate documents in order to respond to litigation and compliance demands, for example. Doing so manually is both time-consuming and prone to human error and therefore electronic systems, such as eDiscovery software, are employed to simplify this important task. Very often, the documents that Deloitte’s Forensics group deals with are image-only files, such as scanned PDFs, JPGs, TIFFs and GIFs. Although humans are able to read the content of these documents on a screen, computers are incapable of reading the content, as it is “trapped” in an unsearchable, non-editable format. To make full use of an eDiscovery’s capabilities documents must first be converted into a searchable format by using optical character recognition (OCR) technology. In order to handle large volumes of files and to automate the OCR process the Forensics group decided to deploy ABBYY Recognition Server.

ABBYY Recognition Server, recently awarded Document Manager Magazine’s Product of the Year, is a centrally-managed OCR service that helps large organisations automate their document capture and archiving processes. The easy-to-use server-based solution makes setup and operation in unattended mode simple, with no need for training. It can also connect with a variety of back-end systems and third-party applications, integrating via a web-service API or a COM-based API. The functionality can also be extended with Scripts and XML tickets.

After comparing multiple products, Deloitte chose recognition Server because of its conversion speed and scalability. Multi-language recognition was another consideration and the solution more than met this criteria with its 190, including major european, Asian and cyrillic-character languages. Throughput was a final consideration, as the team experiences spikes in volume of up to 1000 documents per hour.

With the new seamless solution, documents in the eDiscovery suite that are not searchable are identified and then exported to a watched folder. According to user-defined rules, ABBYY Recognition Server grabs each job, performs OCR on the image files and then creates a fully searchable text file. That file is sent back to the suite so that it can be searched for relevant words or text passages. The solution vastly improves the process for Deloitte, reducing the manual work and the amount needed to search through content to a minimum.