Josh Ballard

Last week, Behavox CEO and Founder Erkin Adylov presented at the Fraud in Financial Services virtual event, alongside executives from the world’s leading banks, insurance firms, and technology vendors.

Erkin’s presentation, titled The Future of Modern Compliance, shone a light on the challenges currently faced by compliance teams and how technology can help them to identify and prevent bad actors within their organizations.

At present, some of the largest banks, hedge funds, and private equity firms are still utilizing legacy compliance monitoring solutions that rely on simple lexicon-based monitoring. When this kind of solution is applied at a firm that has millions or even billions of emails, phone calls, and messages to monitor, the number of alerts generated for the compliance team to assess is staggering.

Erkin explained, “The legacy system is generating anywhere between 2 to 4 million alerts on a monthly basis that a team of three compliance officers has to review. And it’s humanly impossible. It will take them 4,000 days to review all these alerts.”

The solution? “The combination of big data architecture and machine learning that generates 90, 100 times better results in terms of just the sheer volume of alerts generated.”

Behavox Compliance helps to reduce false positives by 90%, freeing the compliance team to use their time to investigate actual instances of misconduct. While this technology enables compliance teams to become more effective at catching bad actors, it’s also helping them to shape their training programs. Erkin continued, “Our client decided that they’re going to use true positives to triangulate their internal compliance training. And what that resulted in is an actual decline in escalations, or true positives, that they’ve seen on a monthly basis.”

The client was able to find patterns in the types of misconduct they found in their communications and then create specific training to combat these issues. “As the client introduced more and more training on the specific topics, what they’ve seen is a declining trend. And what’s really interesting is that they’ve seen a declining trend on these specific topics that have been the subject of training.”

The same client also reported cases of reverse-escalations – employees self-regulating their own behavior before they are caught by their compliance team. “The employees actually are actively reaching out to the compliance team and they’re proactively declaring potential non-compliant behavior. This is what’s called a reverse escalation, where the employee comes to the compliance team proactively and says, ‘we know that you’re using big data technologies, and we know that you use machine learning. We have been to training and we now know that this is something that we’ve done. That was stupid. And we would like to raise our hand and we need your help.’”

Erkin ended the presentation by highlighting the three core components of running a best-in-class compliance surveillance program:

  1. Technology that is able to process billions of data points and provide accurate alerts 
  2. Implementing training related to the instances of misconduct that are found to improve the compliance culture
  3. Employees proactively reporting their own non-compliant behavior having had relevant training