Aurix is not a widely recognized vendor in the speech analytics market even though its technology powers products provided by companies such as CallCopy, KnoahSoft, NewVoiceMedia and OnviSource. In fact its strategy is to provide the engine that these and other partners use in market-facing speech analytics solutions. Its main product, Aurix speech search SDK, is a search engine that searches through call recordings to find specified words, terms and phrases. It uses advanced phonetic search technology rather than the more common technique of large-vocabulary continuous speech recognition (LVCSR) and thus allows users to avoid setting up a library of the words they want to search for. This approach makes it easier to work with different languages and to pick out slang, jargon, product and place names, and other words not normally found in a dictionary.
Consistent with its strategy, Aurix designed the product to be integrated into others, which makes it easy for partners to develop business-related applications to meet their needs and those of their customers. Aurix SDK is also designed to be highly scalable and to work with large volumes of call recordings. Partners can use it to develop applications that identify calls that contain the target content, categorize these based on rules set up by business users and then graphically show call content, call trends, agent performance and customer sentiment. Some of the partners then package these capabilities to deliver “out of the box” solutions to common business issues such as managing first-call resolution, creating more focused and objective agent quality monitoring, training and coaching, identifying breaches of regulations and improving customer satisfaction.
In a recent development Aurix has more publicly released a product with the intriguing name of gopher-it 1.1. This product steps into the space it typically leaves to partners and provides basic analysis and presentation of call content. Version 1.1 can cope with larger volumes of calls and so is more suitable to analyze recordings from large contact centers. Users can set up the product to search for words or combinations of words, and then the system will extract any recorded call records that contain these words. The findings can be displayed in graphs or other basic presentation techniques. This product can meet the needs I find many companies have for speech analytics to gain insights into the content of calls. Aurix has also announced and brought to market a customer contact analytics solution to help provide ready-made application for those that want to provide the insight to others more easily. Companies requiring more complex analysis and more seamlessly packaged business solutions will still need to look for partner solutions.
My research into customer experience management shows that these are early days in the adoption of speech analytics, which I think is partially due to the newness of the products and some vendors overselling what they have as out on the leading edge. I recommend that companies start small and examine what using speech analytics can do for them. Just as with text and social-media analytics, an evaluation will show what products can do and the benefits they can deliver. As one user recently told me, it will also enable companies to understand “what they don’t know” and how these products can help them find out what they need to know and prepare for real business-focused initiatives. In this respect reviewing the gopher-it product will demonstrate what can be done, and companies that want to progress to more advanced solutions could look at the partners of Aurix. Has your company thought about what speech analytics could do for you?
Richard Snow – VP & Research Director