AI risk detection via video, voice, text

Theta Lake was founded in 2017 by two technologists with decades of leadership at Global 100 companies: Devin Redmond, chief executive officer, and Rich Sutton, chief technology officer. The Santa Barbara, Calif.-based firm is a 2019 Gartner Cool Vendor and UC Today’s 2020 Best Compliance Product and Best Security Product award recipient.

Seeing a need for advanced security and compliance systems, the company developed the Theta Lake Compliance Suite, technologies designed to monitor workflows and detect risk in video, voice and unified collaboration systems. These patented AI solutions monitor and flag potentially noncompliant communications across multiple channels and integrated systems, Theta Lake representatives stated. With rising use of video marketing and conferencing, Theta Lake's product suite can help to automatically detect and isolate risks in audible, visual, text and video workflows and archives. The company noted that these complementary assets make it easier for compliance teams to manage MiFID II, FCA, GDPR, FINRA, FFIEC and similar regulatory requirements.

Natively blended technology

Recognizing that legacy email protection systems are ill equipped to handle video, Theta Lake engineers designed tools for newer, collaborative platforms. Instead of repurposing content to make it look more like email, Theta Lake designed the tools to natively blend with different technologies, observed Dan Nadir, chief product officer at Theta Lake. "We started from scratch and built tools that look and work the way you'd expect them to work if you're dealing with video, voice or chat," Nadir said. "We support all major enterprise platforms that people use today, such as WebEx teams, Microsoft Teams, RingCentral for voice, Red Box for voice, Slack for chat and Zoom for video."

Nadir mentioned that enterprise-level data consumption can make regulatory and compliance management especially challenging for organizations. A big company, for example, may lack the resources to effectively monitor 500 daily meetings. Theta Lake applies rules to their systems, using AI to monitor content and separate events that merit attention from broader data sets.

High-definition screening

Nadir additionally noted that financial services and other highly regulated verticals tend to monitor what is said, shared and shown on video conferencing, calls and chats, thus limiting these communications to minimize risk, and Theta Lake's unified interface simplifies compliance risk management by rapidly analyzing text, audio, images, documents, visual content and participants in the video. The AI can automatically detect regulatory and corporate compliance risks across audio and visual content, adeptly mapping items to complicated workflows and directing reviewers to specific locations for further evaluation, he added.

Nadir has observed that reviewers frequently navigate a lot of noise when sorting through data, Nadir stated. Theta Lake can discern between "buy this stock; it will make you rich" and "buy this stock; I guarantee it will make you rich," he said. Alternatively, some legacy systems just pick up the word "guarantee," flagging each time it is said and creating more noise for the reviewer, he added. "Our tools enable reviewers to screen out noise and focus on the things that matter, like complaints, coercive behavior and financial documents on display," Nadir said. "Unlike systems that just read text, our system listens to audio, looks at screens and reads financial documents."

Easy integrations

Theta Lake tools can be integrated with video conferencing, call recording systems and legacy archives, Nadir noted, adding that these capabilities help companies remove compliance inhibitors and leverage video conferencing to improve relationships, increase productivity and reduce travel costs, he added.

Theta Lake's customizable compliance tools are tailored to individual client needs. Clients may choose to screen for offensive language, profanity or weapons on screen, and Theta Lake will adapt its classifiers accordingly, Nadir stated, recalling a recent client who wanted to prescreen documents without reading them. "They asked us to take the first pass and reject a document if something is wrong," he said. "We applied high-risk and medium-risk classifiers that would direct questionable documents to a reviewer who could then reject, approve or escalate it."

Notice to readers: These are archived articles. Contact names or information may be out of date. We regret any inconvenience.