(Image source: Hearsay.)

Hearsay Systems (San Francisco) has introduced the next generation of its enterprise compliance platform, which the vendor says applies predictive technology to reduce false alerts and help compliance teams focus on activities that pose the highest risk, while increasing security coverage and control.

Social Media Compliance Tech with Machine Learning

The latest platform includes the Risk Meter, which the vendor calls the first social media compliance technology that uses machine learning to prioritize highest-risk offenses for supervision teams. Machine learning allows the vendor to take the nearly 100 million interactions it processes every year from more than 150,000 agents and advisors, and use that intelligence to predict the highest-risk items, Hearsay asserts.

Hearsay’s compliance platform was built for the financial services industry and addresses both the industry’s complex regulatory requirements, such as those from FINRA, the SEC and FFIEC, as well as with each firm’s unique company policies, for comprehensive supervision over advisors’ digital actions and communication, Hearsay says.

The newly launched system also addresses the concerns arising from advisors increasingly engaging with consumers in multiple ways including on social, via text and email, compliance teams must monitor more activity across more channels, according to the vendor.

Acting on the Riskiest Advisor Behavior Quickly

“Today, supervision teams are tasked to do more than ever before,” comments Mark Gilbert, CTO, Hearsay. “Hearsay’s powerful combination of machine learning, auto-sampling and auto-remediation allows them to act on the riskiest advisor behavior quickly, while knowing they have broad coverage across their field’s activities. Teams become more productive, with deeper control and expanded coverage over advisor-client communications.”

Compliance and supervision teams supporting thousands of advisors monitor and review enormous amounts of advisor social and digital activity every day, but according to data Hearsay has collected, only approximately 10 percent of activities triggered by lexicon (word-based) monitoring are truly risky. The vendor says that its new compliance platform addresses this problem by applying predictive technology to reduce false alerts and help compliance teams focus on activities that pose the highest risk, while increasing security coverage and control.

Risk Meter uses machine learning to prioritize violation alerts, so teams can address the highest-risk items first, Hearsay says. Additionally, its Intelligent Alert system can catch violations based not just on prohibited words (lexicon), but other, more nuanced data, such as whether similar activities in the past were considered true alerts, thus reducing the amount of false alerts, the vendor sasy. The Risk Meter’s algorithms are trained on regulatory requirements as well as a firm’s own company policies, so that the tool can focus on what is most important to monitor and intercept for the organization.

The vendor describes the solutions new functionality as follows:

• Auto-remediation: Hearsay’s auto-remediation functionality can automatically delete or revert activities performed by an advisor that are a clear violation of policy. For example: If an advisor publishes a Facebook post natively through Facebook, breaking company rules, the Hearsay platform can automatically remediate (e.g., delete) that post.

• Smart Lexicon: Recent Hearsay analysis found that by tuning lexicon rules, the amount of false alerts reduced by 50 percent with minimal impact on the number of true alerts. Hearsay’s smart lexicon feature uses machine learning to help firms more accurately set rules and – by feeding this intelligence into the Risk Meter’s Intelligent Alert system – reduce the number of false alerts. For example: If an advisor publishes a post with the text, “I guarantee tomorrow’s barbecue will be amazing,” and his firm’s policy prohibits use of the term “guarantee,” Hearsay’s algorithms can analyze the post’s full context and other data before deciding whether or not to trigger an alert.

• Flexible Sampling: More flexible sampling provides greater control over “spot checks” on content that was not triggered by lexicon-based monitoring, ensuring potential risks aren’t slipping through the cracks. With 34 percent of Hearsay customers using a version of sampling content as an additional compliance checkpoint, this enhanced functionality gives them more controls – including the ability to create custom rules on frequency and criteria – to sample different types of advisor activity.

Hearsay Introduces New Predictive Social Media Compliance Tools for Financial Services