Global Relay, a leading provider of integrated digital communications and recordkeeping, announces its AI-enabled surveillance solution to support the increased need for communications surveillance to meet evolving regulatory requirements.

“Our surveillance solution alone has been developed by a team of over 300 experts”

The adaptable solution design addresses the limitations experienced by vendors using specialized AI, including poor data quality, constant model training and tuning, dependencies on large training dataset creation for risk or language expansion, tunnel vision, model drift, limited depth of reasoning, and model tie-in. Global Relay’s latest surveillance technology offering allows customers to stay abreast of fast-moving AI developments and deliver future-proof surveillance.

Global Relay Surveillance leverages a five-layered approach, incorporating data standardization and enrichment, transcription and translation, noise reduction, risk identification, and alert management to dramatically reduce false positives, while improving the identification of true positives.

Global Relay’s breadth of data Connectors allows for unmatched data enrichment, delivering structured data from all business-approved communications channels. Near-perfect transcription of voice calls, and translation (across 50+ languages) of all communications ensures that all data is normalized into a single format and language, enabling more effective and efficient analysis.

Leveraging AI, Global Relay Surveillance filters irrelevant data such as spam, disclaimers, or promotional material, to complement traditional methods such as pre-approved content, whitelists, and metadata filters to reduce noise. The remaining data, which has been cleaned and filtered, is then analyzed across a comprehensive set of risks by some of the largest language models available which have a deep understanding of financial markets, lines of business, and regulatory requirements – enabling the identification of compliance and conduct risks.

Global Relay Surveillance demonstrates strong AI capabilities in pattern recognition and contextual understanding, even identifying and understanding challenging evasion techniques such as leetspeak e.g. “5p00f”, removal of spaces e.g. “donttellcompliance”, backward spelling e.g. “drawkcab”, and more. It is able to examine and understand entire email, chat, and voice conversations to make risk-based decisions. For each alert generated, Global Relay Surveillance delivers comprehensive, compliance-based explanations enabling analysts to understand the rationale behind every flagged risk.

Practical alert management tools deliver the fifth layer of surveillance, with user-friendly dashboards, robust alert queues, and intuitive reporting empowering users to manage the end-to-end surveillance process, reducing manual output, saving time, and boosting efficiency.

Hosted within Global Relay’s state-of-the-art green data centers, its AI models process data entirely within the Global Relay environment for robust data security (no third-party cloud vendor or model provider involved). By owning and operating its data centers which incorporate the latest Nvidia technology, Global Relay can run the largest models available, at scale, and at the lowest possible cost.

“Our surveillance solution alone has been developed by a team of over 300 experts,” says Sahar Kayhani, Chief Product Officer for Global Relay. “We’re proud to offer an agile solution with unprecedented scalability for comprehensive communications monitoring that adheres to global AI governance principles and model risk management guidance. This is enabled by a diverse team with a background in trading, regulation, data science and compliance, alongside our product and engineering teams with 25 years experience,”

Global Relay will be exhibiting its surveillance solution for the first time at the XLoD Global Conference in London on November 13-14. A product demonstration will be given at the Global Relay booth, located at #27.