P2P

39 I L T A N E T . O R G Robert Coppola is vice president, legal solutions and strategic growth at QuisLex. At the intersection of law, business, and technology, Coppola is focused on adding value and delivering high quality legal ser vices, solving complex problems by leveraging his unique skills and background for global clients in the most cost-ef ficient manner. Located in NYC, he joined QuisLex after practicing law for over seven years at Quinn Emanuel and Gibson Dunn. the key concepts related to your issue and help build out your issue analysis over time. • Pro tip: Start broad on your issue and narrow your gaze over time. This will make it less likely that any important communications, types of documents or relevant players for your issue are missed. By taking a broader view, you will have a better sense of how the puzzle pieces fit together. • One more pro tip because I like you: If you have a very large data set and you can't find documents related to a particular issue, create the perfect documents for the issue from scratch and feed them into your model. This may help you quickly uncover documents related to your issue that would otherwise stay buried in your data. Compliance Monitoring/Identification of Fraudulent Behavior While it takes investment and foresight, companies can utilize TAR in a programmatic way to ensure their employees are complying with relevant laws, regulations or company standards as well as to uncover any fraudulent or nefarious activities performed by employees. Utilizing TAR for compliance monitoring and/or identification of bad actors/activities may be particularly useful for companies in highly regulated industries. Companies can feed employee communications into previously stabilized TAR models on a weekly, monthly, quarterly or annual basis to ensure compliance as well as to help uncover any concerning or unlawful activities. For example, if you are a financial institution with an active trading arm, you may find it useful to create a TAR model that focuses on your traders' activities. The model can be created and trained based on historical company data as well as public data or created data of both compliant and noncompliant activity. This data can be fed into a model and classified accordingly. Once the model matures, as new data is fed in the most relevant documents related to your focus will bubble up to the high-scored bands for review. This process can help companies stay on top of potential issues before they fully arise. Plus, regulators will look kindly on your efforts if an investigation does come up in the future. Looking to the future For TAR to continue being useful in tackling the above instances and others in the future, it will need investment in continuous improvement. The data landscape is always changing. Smartphones, social media, chat/messaging systems and myriad collaboration platforms are all becoming common within organizations for communication today. These new technologies generate large amounts of complex data that must be analyzed to understand the totality of your data picture. Current TAR models are not built for these new types of data – these unique data types were not in scope when TAR first started being utilized. While it has not been the focus of this article, the solution to this challenge requires evolving and updating TAR workflows to include more advanced AI technology with trained experts and linguistics to help wield them. For TAR to remain effective in an increasingly complex modern data environment, it is necessary to create and innovate tools for the TAR process that leverage more advanced subsets of AI, such as deep learning and natural language processing. By doing so, TAR will continue to a be an important weapon in our arsenal to wield against eDiscovery workflows and provide better outcomes for our clients in the most cost-efficient manner achievable. ILTA

• Cover