eDiscovery Leaders Live: Charissa Bass of KPMG
Charissa Bass, Director, Analytics and Machine Learning at KPMG LLP, joins George Socha, Senior Vice President of Brand Awareness at Reveal, for ACEDS #eDiscoveryLeadersLive.
Charissa Bass is a Director in KPMG’s Forensic practice at the Charlotte office. For over a decade, she has provided Data Analytics consultation in eDiscovery for Fortune 100 and AmLaw 200 clientele, using innovative, multifaceted approaches that emphasize machine learning, to overcome challenges of Big Data in large-scale litigation, regulatory investigations, corporate information management, and time-sensitive internal investigations.
Charissa began the session talking about her background, her role at KPMG, and the early days of TAR. She then used the podcast to announce KPMG’s new partnership with Reveal. She discussed why they chose Reveal, testing and evaluation they did, and what they hope to do with the capabilities Reveal offers, highlighting the AI model library and the workflows they are building with Reveal. Charissa also used the podcast to talk about a new service that KPMG is offering, RCA or Rapid Communications Analysis – what it is and what it will provide to clients. She shifted to a broader discussion about analytics, including active learning, covering adoption levels, what they are doing with these capabilities, and how clients are reacting. Charissa closed with thoughts on what she is looking forward to next.
Key Highlights
- [1:57] Charissa’s background and role at KPMG.
- [3:14] How she got into analytics and machine learning.
- [4:13] The early days of TAR.
- [9:19] Her day-to-day role.
- [7:34] Announced KPMG’s new partnership with Reveal.
- [8:13] Why they chose Reveal.
- [8:53] Testing and evaluating Reveal.
- [10:10] What they most look forward to doing with this new set of capabilities.
- [10:41] Client reactions.
- [11:17] Building their own AI model library.
- [13:01] What they hope to accomplish with their AI models.
- [13:25] Developing workflows.
- [13:34] A new service offering: RCA, or Rapid Communications Analysis.
- [15:03] Educating clients about how much – or little – data is needed to do meaningful analytics.
- [16:40] Using customer AI models and in-depth analytics with RCA.
- [18:10] The art of telling me something I don’t know.
- [18:26] What KPMG’s new service offering will involve: complete analysis, including communications and sentiment.
- [18:40] Thoughts on sentiment analysis.
- [19:15] Thoughts on the shift from traditional TAR to newer forms and more creative uses of active learning.
- [20:16] The many ways KPMG is using active learning.
- [22:21] A shift toward clients asking for more powerful tools.
- [24:00] The inevitable increase is AI adoption.
- [25:07] Who they work with.
- [25:32] What clients hope KPMG will be able to do for them.
- [27:31] What she is looking forward to next.
Key Quotes
- “We believe that Reveal is the future and will have the market share.”
- “We were really honored to be part of the process of doing the testing and becoming an early partner so that way we can help shape the functionality and the capabilities going forward.”
- “We’ve jumped on demoing [Reveal] already to our clients and we’ve gotten nothing but great feedback and we already have live client data in there. I did a demo for a client recently and they were blown away.”
- “We actually just came out with an awesome new service offering that I wanted to mention here…. It’s called RCA. It stands for Rapid Communications Analysis…. It is similar to an early case assessment, but it’s just focused on communications…, the ability to find insights in the data, telling the story of what happened. These will be your internal investigation type scenarios where … and a company needs to act very swiftly in order to stop the bleeding and to defend their reputational risk and get prepared for possible litigation.”
- “The analysis is a bit of an art. The human is still needed. You can get lost in the tools, you can get lost in a communications graph or a cluster wheel. So you really have to [learn how to use the tools] and to learn from the client what it is that they’re looking for, what they know so far, and then you can help them with the tools such as concept searching. You can help them identify the unknown unknowns, things they did not even know to look for.”
- “We pretty much apply active learning on every matter. Even if we are not using it to cull documents from the production, we will use it to maximize efficiencies for our clients.”
- “We are so ecstatic to get our hands on Reveal. It’s the only tool that I know about right now that has the sentiment analysis and the anomaly detection….”