Fonetic has launched new enhancements to its communications surveillance and trade reconstruction platform Trade Comms Suite.
These updates will allow financial institutions to adopt voice and communication surveillance quicker while taking advantage of the most accurate voice surveillance platform currently available.
Among the additions are improved precision to financial language models and speech signal treatment. The new surveillance engine, which has been evaluated against the leading voice surveillance vendors in numerous proofs of concepts, has surpassed the nearest competitors by 15 per cent according to the data received. Apart from offering high accuracy in English, Fonetic is available in more than 20 languages across the world and has been deployed on trading floors since 2008.
Fonetic’s capabilities in Machine Learning (ML) have also been enhanced. The ML engine, which includes entity detection algorithms, has been tested against the Snips corpus and showed better results on this corpus than both Snips, Google and Microsoft. Fonetic Entity Detection is not only restricted to the entities of this corpus but focuses its efforts on financial-specific entities; enabling the detection of information as specific as types of swaps and term dates, among others. The accuracy on these reaches 90 per cent (F1).
Twenty new Machine Learning classes are now trained and ready to use, including intent detection and communication classification. This enables the compliance function to not only greater reduce their false positives in surveillance but also respond to regulatory requirements on Trade Reconstruction.
The improved solution, already rolled out to some current clients, combines faster communication processing with more resilient functionality, allowing Financial Institutions to reduce their infrastructure costs, both in cloud and on premise.
CEO and Founder Juan Manuel Soto, says: “Innovation for customers is the foundation of how we function. With this release, by improving how we handle communication data, we can deliver faster and more accurate results to our clients; helping them react quicker to incidents they come up against.”