Let Us Guide You

Select Industry
Select Business Problem
Find Solutions

Knowledge Center Nexidia Approach

The Technology Behind the Magic

Nexidia has always invested in original speech technology research and invented the process of rapidly searching audio known as phonetic indexing.  Today, Nexidia employs Neural Phonetic Speech Analytics™, which combines the strength of Automatic Speech Recognition (ASR) and phonetic indexing in a way that offers the most scalable, accurate and flexible solution available.  

Phrase Recognition
Nexidia technology is based on two parts, phrase recognition and phonetic indexing and search.  Nexidia Phrase Recognition is a process which produces both word-level transcription of the audio and customer sentiment scores using advanced neural networks and sophisticated language models.  A model of the acoustics is used, based on real-world conversational audio, as is a model of words and likely word sequences.  Seeding this process is customer text, such as web-crawls to identify any specific terms that we should be expecting, to reduce the likelihood of not having all words in the dictionary.

The text produced can be analyzed to facilitate the rapid discovery of interaction content, uncover emerging topics, quantify related phrases and categorize interactions by topic based on business logic. 

Phonetic Search
Neural Phonetic Speech Analytics™ also generates a time-aligned phonetic index.  This index is based on phonemes, the distinct sounds that make up language.  Because phonemes are simply uttered sounds, the indexing is not affected by factors such as background noise, languages, dialects or speaking styles. 

The phonetic index can be searched directly on words or phrases or using special operators such as Boolean strings or time-based proximity to other content. Nexidia’s proprietary search engine identifies and matches the phonetic equivalent of the search string and returns relevancy-ranked results. When words or business logic need to be changed, the system can rapidly re-query the index without having to re-process the audio. The result is a process that not only creates the truest representation of spoken audio, but also enables the fastest, most accurate access to the information contained within the audio files.

Whether used to improve the tagging of audio for syndication, to perform ad hoc searching for audio discovery and evidence, or for analysis and reporting on large volumes of customer call data, Nexidia delivers advantages no other technology can match.

Let the power of Nexidia's technology begin working for you today.

Contact us to learn how