Taking advantage of high-powered computers and servers in The Cloud, long-time speech and text analytics specialist Nexidia is expanding the footprint of its core product well beyond the confines of the contact center. Touting the power of Neural Phonetic Speech Analytics™ to support Large Vocabulary Continuous Speech Recognition (LVCSR), Release 11.o of its Interaction Analytics Platform supports levels of precision in word accuracy enabling the sort of “deeper recognition” of meaning that business decision makers seek when looking for insights that can shape company strategic planning and marketing decisions.
The new architecture and approach promises to scale massively. One of its communications clients, for instance, is already capturing and analyzing 70,000 to 100,000 hours of audio each day and storing it for over a year for future analysis. As Nexidia explains, using LCVSR “underneath” neural networks enables business decisionmakers to make discoveries at the business unit level across 100% of the captured audio in order to support “metrics-based management.” The new release also moves toward real-time analytics, with support of up to 44 languages, assuming the purchase of additional language packs. The approach also enables business managers to merge text input from social media posts and tweets as well.
The benefit for large companies with high volumes of interactions is clear. Nexidia tells us that smaller companies can also benefit when the deployment is made in a multi-tenanted architecture. Medium-sized businesses will be able to perform the same sorts of queries in real time at competitive prices.