Peer to Peer Magazine

P E E R T O P E E R : I L T A ' S Q U A R T E R L Y M A G A Z I N E | S U M M E R 2 0 1 9 15 evolving system that enables continuous improvement over time. Machine learning algorithms rapidly detect patterns in large bodies of data. As these algorithms are repeatedly exposed to thousands or millions of specific data points and progressively larger amounts of related data – and as they are exposed to feedback and corrections from informed users – they begin to make increasingly accurate inferences, steadily and iteratively refining our understanding the data. Through this process, machine learning algorithms are able to make increasingly accurate predictions about the kind of information a new data set is likely to contain and provide insights that would be impossible to access through other means. Multi-Matter Use Cases Machine learning and other analytic technologies are most powerful when they are applied to large volumes of complex information. We have only begun to tap these technologies in the legal domain, but one of the most promising current scenarios is their application to data from multiple matters or to an organization's entire legal portfolio. Many of us have already observed that targeted use of predictive coding can produce big cost savings even in the context of a single legal matter, but analytics can take us much further, making it easy, fast and practical to leverage insights about the nature of the data derived from one matter and apply them to other matters. For instance, privileged documents from one matter are highly likely to be privileged documents in subsequent matters. Finding such documents by keyword searching or other means is relatively expensive and time-consuming the first time around. Analytics can dramatically speed up the process and produce more accurate results in new matters by applying what it has "learned" in previous ones. Similarly, analytics can be used to quickly isolate "hot" or especially relevant documents based on results Integrated Analytics Is a Game- Changer But change is in the air. In the past few years, analytics-based applications like TAR and predictive coding – which leverage machine learning and predictive modeling – have begun to have a major impact in the cost-heavy area of document review, particularly in high-volume cases. That development alone is significant. Even more important, however, are recent efforts to deploy artificial intelligence and analytics technologies across a much larger segment of the litigation workflow and across multiple legal matters, a development that will soon prove to be transformative. When applied thoughtfully, these innovations will help turn the legal department into a model organization that consistently and effectively manages profit and loss, and repeatedly demonstrates its value as a proactive and strategic player in the boardroom. The emergence of analytics as a game-changer for legal departments is occurring in the context of broader technological innovations – including SaaS and cloud computing, certainly, but also agile development methodolo and platform extensibility, which encourage the development of systems that can rapidly adapt to changing requirements. Within this more flexible and scalable computing environment, technologies like machine learning, predictive modeling, data analytics and data visualization are already helping GCs identify and continually monitor key performance metrics for their organization at every point in the litigation process so they can granularly understand, predict and budget for eDiscovery and litigation spending. Machine Learning Enables Proactive Decision-Making The proactive potential of analytics is perhaps best embodied by machine learning, which transforms essentially static technolo focused on automating data- based processes into a dynamic, constantly from previous matters that are focused on comparable legal issues. In internal investigations, analytics is being used to quickly identify code words, project names, custodians and other data points from previous investigations that are statistically likely to be relevant in the new matter. Analytics confers a huge advantage on legal teams by surfacing important evidence (and filtering out "noise") much sooner in the litigation process, enabling earlier – and more facts-based – decisions about whether to settle or which legal strate to pursue. This has obvious implications for early case analysis, and it has the potential to save legal department lots of money. Analytics- derived metrics and processes can also be used to inform work on multiple matters simultaneously in real time. A good basic example is document tagging in review: Decisions from one matter can instantly be applied to another, similar matter that is being litigated at the same time. The application of analytic tools across multiple matters is giving GCs and legal teams invaluable new information that's allowing them make detailed, accurate projections during the eDiscovery phases of the litigation workflow. The more we expose the legal portfolio to analytics, the more we can precisely we can anticipate data volume, the number of individual documents, the document types, the number of custodians and the number of reviewers a new matter is likely to involve. When we aggregate this information over time, we begin to get data-based answers to fundamental questions, including questions about the operational side of the litigation workflow. Based on past performance in similar matters, how much can we expect to spend on document review in this particular matter? How many reviewers will we need to stay on schedule, and which reviewers are likely to be the most accurate and cost-effective to use in the current case? Which are not? Which of our outside firms are making the most efficient T H E R E ' S M O R E O N L I N E ! Want to learn more? You're in luck! Click the link below to listen to the audio that corresponds to this article! L I S T E N T O P O D C A S T

• Cover