Digital Reasoning, a leader in Artificial Intelligence (AI) that understands human intentions and behaviors, has launched a new, AI-enabled voice analytics solution for voice communications monitoring in banks. Developed with extensive input from the firm’s industry partners and customers, it delivers market-leading transcription accuracy of financial domain audio data and seamless workflows that enable efficient scaling of audio monitoring coverage. Allied with Digital Reasoning’s award-winning Conduct Surveillance solution, it enables banks to achieve integrated surveillance that covers both voice and text communications data.

“Banks are under increasing regulatory pressure to improve monitoring of voice channels but need a solution that enables surveillance scope to be expanded with the same staffing,” explained Brandon Carl, Vice President of Product Management at Digital Reasoning. “Our proprietary solution uses advanced AI to deliver performance and efficiency. It improves signal-to-noise ratios and enhances the quality of alerts, and adds smart investigation tools that zero in on relevant audio data.”

Key solution benefits:

• Integrated voice and text analytics reduce analytics silos, enable more efficient workflows, and help banks make progress towards holistic surveillance.
• Best-in-class transcription engine, tuned to conduct surveillance applications, produces more accurate transcripts that improve the identification of risk and reduce false positives.
• Comes with pretrained and domain-specific language models, which are fully customizable using customer data to achieve unparalleled accuracy.
• Scalable to enterprise populations and multiple audio channels, adapts to English accents, and features pinpointing of alerted phrases to speed investigations.

Banking environments - such as trading floors - typically generate noisy, fast-paced audio data that is peppered with domain-specific language. Digital Reasoning was able to draw on its deep relationships with leading banks and its extensive experience in conduct risk monitoring to develop a solution that offers market leading accuracy when tasked with analyzing live and archived financial domain audio data. It “sounds out” each character to produce precision transcriptions, allowing complete conversations to be captured for semantic and contextual analysis, resulting in improved risk identification and reduced false positives. 

The new solution packages advanced audio analytics capabilities with workflows and tooling that enable surveillance teams to dramatically upscale audio monitoring coverage without adding to numbers or workloads. Integration with Digital Reasoning’s proven Conduct Surveillance allows comparable misconduct models to be applied across voice and text through a single workflow. This reduces analytics silos, improves efficiency, and gives confidence that a bank’s improved voice surveillance capability will hit the same regulatory targets established for text.