The eDiscovery Paradigm Shift

Tuesday, April 3, 2012

Zoom to Predictive Coding

Predictive coding has captured the imagination of the eDiscovery market.    As a result, my daily conversations with members of the legal departments of the global 2000 and the eDiscovery professionals within the major law firms now routinely include questions about how predictive coding works, what predictive coding technologies are available and which predictive coding vendors I recommend. Therefore, in an effort to better serve my curious client base, I caught up with Warwick Sharp, co-founder and Vice President of Marketing and Business Development for Equivio, to learn more about their predictive coding platform.

Founded in 2004 in Israel, Equivio began by developing and marketing near-duping and email thread management software. Its acquired a wide variety of clients, including hundreds of law firms, corporations, government entities and consulting organizations throughout the world. Its client base includes the Department of Justice, the Federal Trade Commission, KPMG and Deloitte. In addition, Equivio technology has also been a favorite integration partner among other litigation software vendors such as kCura and many of the well known litigation service providers.

In early 2009, Equivio made the gutsy decision to enter the unknown and highly under appreciated predictive coding wilderness with Relevance, its standalone predictive coding platform. Along with a handful of other predictive coding pioneers such as Recommind, Orcatec and Xerox, Equivio set out to convince the legal community that this new software with its complex mathematical algorithms and confusing statistical models could do a better (i.e. more statistically significant) job of identifying relevant documents than human reviewers. And, although I haven’t asked Equivio about the early financial returns on this bet, I would suspect that the initial missionary marketing efforts were tough and didn’t produce a financial return on their investment. However, Equivio didn’t give up. And, with the recent landmark court decision by Southern District of New York Magistrate Judge Andrew Peck on the Da Silva Moore case opening the flood gates for the legal tolerance for computer assisted review (aka predictive coding), they are now well positioned as one of the few legacy players in the predictive coding market.

On January 23, 2012 Equivio launched Zoom, a single integrated platform for predictive coding and analytics. As reported in their initial press release, Zoom combines Equivio's best-of-breed near-duplicates, email threads and Relevance components together with data import and export, early case assessment and enriched analytics. These components are seamlessly integrated on a unified web-based platform for easy access and use. This past week, I had a chance to spend some time with Warwick and see Zoom in action.

My first impression of Zoom is that it has an extremely intuitive and attractive web-based user interface that provides an easy to understand workflow management system to lead users through the process of collection, processing, analytics, first pass review with predictive coding and export. Support for collection, processing and initial culling appear to meet or maybe even exceed industry requirements to prepare data for initial review.

From a more detailed standpoint, Zoom enables users to extract text and metadata from raw ingest and generates data profiles for Early Case Assessment (ECA).  It then enables users to subject the data to multi-layered analysis, including de-duping, near-duping, email thread management and language detection. Zoom's search environment also supports keyword analysis and metadata faceting.

From a predictive coding perspective, Zoom enables litigators to efficiently “train the system” to identify and assess documents for responsiveness and privilege with little or no knowledge beyond standard online document review best practices.  The platform displays the complex, yet required, predictive coding statistics with easy to understand graphics and even provides users with the cost of review for the current level of statistical significance and the cost projections for increasing that percentage. Finally, Zoom exports native files, full text and metadata to the user’s review platform of choice. Please note that I have not tested the viability of Zoom’s predictive coding accuracy and therefore, for the purpose of this initial review, I am taking Equivio’s word that it meets or exceeds expectations for computer assisted document identification and that the results would hold up in court.

I have designed, tested and used literally hundreds of Web-based applications over the years and I would rate Zoom among the best for easily guiding the user through a fairly complex process and providing an initial positive experience.

And although Warwick was reluctant to position Zoom as an Early Case Assessment (ECA) tool, I believe that Zoom does in fact have the foundational architecture, features and a workflow that can mature into a formidable competitor in the ECA market. Whether or not Zoom fulfills this destiny will of course be up to Equivio and how they decide to position and market their new product.
Given my position as an impartial industry analyst with the responsibility of being a trusted advisor to my clients and because there are other noteworthy predictive coding platforms on the market from vendors such as Recommind, Orcatec and Xerox, I am reticent to declare Zoom as the leading predictive coding platform in the industry.

And, given the fact that eDiscovery Solutions Group, in partnership with DCIG, will be including Equivio in its 2012 Early Case Assessment Interactive Buyer’s Guide available in May 2012 and the 2012 Predictive Coding Interactive Buyers Guide available later this summer, buyers will have all of the objective information that they require to make their own decisions.

However, after my initial review, Zoom should definitely be on any buyer’s short list of next generation eDiscovery platforms as it definitely provides the culling and analytics required for serious Early Case Assessment (ECA) and integrates an attractive and competent predictive coding component.

For a more detailed overview of Zoom, you can contact me at: productbriefings@ediscoverysolutionsgroup.com.

For more information on Equivio and Zoom, please visit: http://www.equivio.com/.
Click Here to join the International Association of Predictive Coding

Labels: eDiscovery, Equivio, Litigation, Orcatec, Predictive Coding, Recommind, Saas, Software-as-a-Service, Xerox

Tuesday, August 30, 2011

eDiscovery from Magic to Mainstream Technology

Is the current state of eDiscovery sufficiently advanced from a technology standpoint that it is indishtinguishable from magic? Probably, but that isn't necessarily a bad thing.

It wasn't that many years ago that only a hand full of ligitation technology "geeks" knew what eDiscovery (Electronic Discovery or E-Discovery) even meant. In fact, it really wasn't that many years ago that Information Technology (IT) had little or nothing to do with supporting the fine art of litigation. Don't get me wrong, there were plenty of disagreements about the ownership and use of intellectual property and associated technology that were adjudicated in a court of law long before anyone had ever heard of eDiscovery. There just wasn't much technology, if any, used to support the litigation process.

Litigation has historically been the dominion of lawyers with visions of men in white wigs and "Sherlock Holmes type" investigators gathering and presenting evidence under a set of rules that were completely foreign to all those except members of a club called the bar. Perrry Mason dragged litigation and the "experience within the courtroom" into the post World War II era and made it a bit more glamerous and attractive. But, there still wasn't much technology involved in the litigation process.

Typewriters and the Gold Era of Litigation
The introduction of the IBM Selectric typewriter in the 1960's revolutionized business around the world. And, in conjunction with the first commercially feasible copy technology in the 1960's, thanks to Xerox, the volume of paper expanded exponentially (Please note that I am not sure that anyone except a few math geeks used the word exponential at that time). I think that the rule of thumb for most enterprises at this time was that if 1 copy was good, 20 copies was great! So, with the proliferation of paper copies of everything, litigation had to accommodate this new paradigm of massive volumes of paper evidence. And, hence the introduction of the cardboard legal box and what we all now fondly refer to as "document review". Since most evidence was now on a piece of paper somewhere, every document had to be copied multiple times so that teams of lawyers (on both sides of the case) could review every single document (usually multiple times) and copies of all of this paper also had to sent to the courts. Some in the copy and document review business probably call this the "golden era of litigation" as there was millions of dollars spent on just copying and reviewing documents before the lawyers could even begin to get ready for trial. Under any circumstances, the amount of potential evidence exploded and therefore litigation became very expensive during this time.

Scanning and Optical Character Recognition (OCR)Sad for many but true, the paper only era was actually short lived in the overall timeline of present day litigation practices. By the 1980's, the Professional Computer (PC) with word processing was replacing the typewriter and a new paradigm shift was underway with the advent of Electronically Stored Information (ESI). However, there was still lots of paper left over from the previous "owners" and still a fair of amount of paper being printed (and not stored) in the present paradigm. So, beginnig in the 1980's and becomming prevalent in the 1990's commercially feasible scanning technology along with Optical Character Recognition (OCR) software hit the litigation technololgy market and now paper could be scanned and converted into searchable ESI. The litigation service provider market emerged almost over night turning many with little more than a scanner in a warehouse into multi-million a year businesses. With the advent of the scanner and OCR, the litigation cost per page went down. But, with the accellerating volume of ESI, the overall cost of litigation continued to go up. Unfortunately, the amout of information and the costs were about to go up even more.

eMail and the Internet
Even though PC's were prevelant throughout the enterprise in the 1980's, it was until the late 1980's and early 1990's with the advent of the commercial World Wide Web (Internet) and the subsquent explosion of email, that ESI really began to become the driver of the next big paradigm shift in eDiscovery.

Litigation Software and MagicWith the explosion of ESI in the 1990's, the technologists began to realize that there was a very lucerative and, in many cases, captive audiance (i.e. you can't hide from litigation) that needed software to drive ESI through the litigation process, find the specific data required, increase productivity and reduce costs (please note that I am not convined that the cost reduction criteria was real high on the list of benefits of the early litigation software vendors). And therefore, almost overnight, a whole new industry was born and what I am referring to as the "magic" began.

In an industry that just 40 years earlier was barely dealing with paper, the introduction of ESI and the need collect, processing, find/search and review data was probably not that much different to many litigators than fire was to the early cavemen (no offense meant to cavemen). The legal process (i.e. the courtroom and associated rules) had not changed. However, the entire playing field of where most evidence came from and how it had to be handled and how it had to be presented was totally diferent. For all practical purposed, it was magic to most in the legal industry. Remember that sufficiently advanced technology is indishtinguishable from magic!

Social Networking and Mobile Devices
Today, with the dawn of cloud computing, the era of email comming to an end and the exponential explosion of social networking platforms such as facebook, Twitter and LinkedIn (just to name a few of the big players) on mobile devices, the true era of ESI and the subsequent birth of true eDiscovery has really just begun.

To put this into perspective, there 30 billion pieces of content (e.g., links, photos, notes) shared on Facebook each month and there were 25 billion tweets sent on Twitter in 2010. I could go on and on with the staggering statistics. However, I think that we all get the point. Social media has changed the world and as a result, the practice of eDiscovery will never be the same.

All of this being said, we are making tremendous progress moving from magic to mainstream technology.

Changes to the FRCP

On December 1, 2006, the way that litigants in federal civil lawsuits conduct discovery changed to reflect the increasing prevalence and relevance of ESI in legal proceedings. Prior to this change, many courts had been grappling with issues surrounding the treatment of electronic records discovery. The new amendments to the Federal Rules of Civil Procedure (FRCP) in essence made it more difficult to use the fact that information is held in electronic form as a defense to fulfilling discovery requests. The requirement that is implicit within the new eDiscovery amendments is that litigators must now plan ahead of time to better organize and manage their vast stores of information.

Pertinent Case Law

Zubulake v. UBS Warburg is a case heard between 2003 and 2005 in the United States District Court for the Southern District of New York. Judge Shira Scheindlin, presiding over the case, issued a series of groundbreaking opinions in the field of electronic discovery. Plaintiff Laura Zubulake filed suit against her former employer UBS, alleging gender discrimination, failure to promote, and retaliation. Judge Shira Sheindlin's rulings comprise some of the most often cited in the area of electronic discovery, and were made prior to the 2006 amendments to the Federal Rules of Civil Procedure. The relevant opinions in the field are known as Zubulake I... and Judge Shiendlin is now touring the country/world speaking on the merits of this case and how it has changed eDiscovery.

And, as time goes on, there are more and more cases everyday defining this new thing that we call eDiscovery and the associated technology.

Mainstream TechnologyWith the advent of "big data analytics", "cloud computing, "fatter/faster Internet pipes" and Software-as-a-Service (SaaS), the magic is wearing off and litigation and eDiscovery technology are headed towards becoming mainstream. Workflow models and associated methodologies such as the Electronic Data Reference Model (EDRM) along with standard data exchange formats and standard federated datastore formats, are all moving the industry to a point where it will normalize, become somewhat commoditized and costs will continue to come down. However, that doesn't mean the industry will not progress with the magic comming back.

As an example, the industry is in the middle of trying to figure out how to deal with all of the ESI in the cloud and at the same time wanting to move its tools and platforms to the cloud. So, some are scared with with the concept of multi-tennant applications and whether or not they should go with private, public or some hybrid cloud infrastructure. Others just think that this cloud thing will pass like some bad storm.

And, of evern more interest from a pure progressive software standpint, the current discussion/debate about "magic" in the industry is in regards to "predictive coding". With vendor wars over patents and approaches and courtroom discussions/hearing about methodologies, algorithms and statistical significance, it is really heating up and providing some real entertainment for the peanut gallery.

As a litigator, you are probably in the wrong business if you don't love the continual progression of magic and the evolution to mainstream technology. As a technologist interested in progressing information governance as whole, it doesn't get much better than the rapid succession of paradigm shifts that have occurred in eDiscvoery in our life times. I love it!!

And, in the end, the real winners will be the legal system and those of us who have to deal with it. As a result of all of this magic, litigation will be faster, more accurate and cost less (At least theoretically).

Labels: Copier, Costs, Early Case Assessment, eDiscovery, FRCP, IBM, OCR, Paper, Scanning, Typewritter, Xerox

This Page

The eDiscovery Paradigm Shift

Tuesday, April 3, 2012

Zoom to Predictive Coding

Tuesday, August 30, 2011

eDiscovery from Magic to Mainstream Technology

Contributors

Links

Previous Posts

Archives