Compensation Investigation: A Reveal Case Study

Background: internal audit

A financial services client needed to perform an internal audit to determine who, if anyone, approved certain compensation agreements. Email communications and electronically stored documents were to form the main corpus. They needed a way through the data to piece together a narrative and establish a timeline of events. This also included identifying what other key individuals within the company were aware of the agreements, and to establish a timeline of events.

Challenges: key individuals no longer available

There were a number of challenges to overcome for the client to succeed. Initially, there were 4.1 million documents collected for review after using custodian and date filtering parameters. Keywords and domains further culled the data to 450,000 documents to review. Given the nature of the investigation and the size of the document population, there was not enough internal bandwidth or time to manually review the data. Furthermore, the key individuals that approved compensation agreements also no longer worked with the firm.

“By integrating and leveraging the individual strengths of the tools in Brainspace, we were able to limit
review to only .1% of the original data population while satisfying our obligation to the internal auditor.”

Solution: how did they use the tech?

The client provided four example documents which formed the key language that was used to run several Concept Searches. After these initial results were reviewed, the client used the Cluster Wheel to further identify similar material to the key documents they had just identified.

There were several Focuses for the team to review. These contained communications between discrete individuals during the relevant time period. The client also used Thread Analysis to better understand key, complex email threads that were in circulation for several months. These proved to be pivotal in understanding who knew what and when.

Supervised Learning was continuously run in the background. This ensured that similar documents to those being marked relevant by the other workflows, would be prioritised for review. Quality assurance checks were applied over the data to ensure all key documents had been captured and reviewed. Finally, the client used the Communications and Conversations visualizations on their key document set to quickly and easily present their findings to their Board.

Results

By integrating and leveraging the individual strengths of the tools, the client could limit the review to 5,000 documents, satisfy its obligation to their auditor, and feel confident in their final report.

See this link for more information on the AI power behind Reveal. For more information on this article please see Reveal Resources.

Visit the Siera Data on LinkedIn

Back to Resources Tab

alt View PDF | Download PDF
Share with friends
alt
alt