Senior lawyer discussing with female coworkers at table seen through glass in library

Case Study: Prominent Southeast Law Firm

Ricoh's eDiscovery Managed Review Services Team Uses AI-Driven Approach to Meet a Quick-Turn Deadline While Dramatically Reducing a Law Firm's Document Review Costs 

About the customer

In practice for almost two decades, a Southeast law firm has a centralized office, nearly 100 lawyers working in office or remotely across the U.S., and dozens of practice areas. The firm serves more than 15 different industries and many of their attorneys have been named Super Lawyers, Rising Stars and Legal Elite. Since its inception, the firm has been dedicated to diversity and more recently, to social justice causes.

Kudos to Ricoh for their review team and AI-enabled technology to help get a fantastic settlement for our client.

Law Firm Partner

Challenge

  • Quick turnaround to analyze 800,000+ electronic records

  • Growing client reluctance toward costly linear document review

  • Complexity of analyzing many document types

  • Lack of automation at the firm to speed document review

As with most high-profile law firms, a small percentage of cases proceed to litigation. In one such case, the firm faced a quick deadline of just over a month to produce case data. They didn't have the capacity to tackle the job in-house, but they didn't want to spend hundreds of thousands of dollars on outsourced eDiscovery – especially on a case that wouldn't yield a large award if they settled or secured a win in court.

The firm's client was a commercial property enterprise and they were waging a construction defect case against a contractor. The contractor pulled sub-contractors into the case, adding to the volume of documents that had to be analyzed. In turn, the defendant filed a counterclaim, which required both sides of the case to enter a fast-tracked eDiscovery period.

When the data arrived at the firm, there were more than 800,000 electronic records that had to be analyzed. The data consisted of many different  document types from construction drawings and engineering reports to email communications and progress reports.

Results

  • Saved $180,000 in potential Managed Review costs

  • Reduced documents for review from 198,000+ to 16,000

  • Reviewed and produced documents in 2 weeks

  • Quick, favorable settlement for law firm's client

Reviewing the entire data population of more than 800,000 documents would have potentially cost the law firm nearly $650,000. Using our Technology Assisted Review (TAR) workflow, our eDiscovery Managed Review team was able to reduce the number of documents for review to just 16,000 – cutting the firm's review costs by nearly 98%.

Ricoh's eDiscovery expertise and technology also expedited the review and enabled production of all relevant documents in just 2 weeks after the review began. And the firm was able to quickly reach a favorable settlement for their client.

This quick, cost-efficient outcome thrilled the law firm. So much so, they have been referring Ricoh's eDiscovery Managed Review team to other attorneys at the firm. They know Ricoh has a cutting-edge, proven process that dramatically reduces costs and produces data that will stand up in court.

How We Did It

  • Ricoh eDiscovery Managed Review Services

  • Technology Assisted Review (TAR) workflows

  • Filter searches, DeNISTing and email threading

  • Use of artificial intelligence (AI) and active learning technology

With such a high-pressure, high-speed turnaround, the firm turned to our eDiscovery Review team to analyze all of the ESI for the case. A trusted partner for over two years, the team was asked to do this as quickly and affordably as possible. Key to this was using our Technology Assisted Review (TAR) workflows to efficiently organize, search and reduce data sets to only the most relevant, responsive documents.

Without having to manually look at a single document, we ran keyword, domain and date filter searches using our processing tool to reduce the number of documents that needed to be reviewed. We also performed DeNISTing to remove file types that were unlikely to have evidentiary value. By running a secondary DeNIST list that contains even more program files, we are typically able to remove an additional 10% to 20% of documents. Through this culling process, we cut the total number of documents that needed to be reviewed to 407,000 – a reduction of nearly half. But we weren't finished.

Next, we conducted email threading – a process that reduces long email threads to just a single email to review – and further reduced the number of documents for review to 198,000. From there, we used artificial intelligence to perform active learning and review this batch of documents for responsive and non-responsive categorization – a process that learns from our coding as we reviewed documents. Our tool automatically learned and assigned a rank based on an algorithm as to whether a document was relevant to the case or not. On a scale of 1-100, our team could see that any document that ranked below 60 was unlikely to be responsive. To be certain, we also performed Elusion testing to measure the validity of the results. Through this active learning process, we ultimately cut the number of documents that had to be reviewed to 16,000.

Download PDF