In the contact center will continue to support Aurix’s existing customers and channel partners. It is a product that has proven its mettle in the classic context of deploying analytics on an existing corpus of stored voice files to support agent training and workforce optimization. It is also the root of delivering “relevant screen pops” based on the context of the call and “nuance detection” aimed at both agent and caller.

But that’s only the beginning. The acquisition was driven by the Emerging Technologies business unit, under VP Chris McGugan. He sees the Aurix core engine and IP as foundational technology that can be spliced into a variety of solutions across the entire enterprise. The same attributes that make it possible for the Aurix phonetic speech search engine to monitoring up to 100% of recorded calls makes it possible to do the sort of “near real-time” speech processing to tag the content of the streamed audio associated with video content or collaboration platforms. This, in turn, enables real-time identification, searching and audio mining.

Avaya’s new in-house approach to speech analytics turns it into the connective tissue that can help its enterprise customers connect the dots between other “Big Data” analytics initiatives. It’s no coincidence that only yesterday database and CRM giant Oracle purchased Endeca, a small New England-based company that specializes in analytics and pattern recognition across large repositories of both structured and unstructured data. The difference is that Endeca conducts its analysis of text and other content on the Web and e-commerce sites. Note the lack of speech analytics. In the world of multi-channel, multi-modal and real-time commerce, Avaya recognizes that treating content and meta-data that starts as spoken words may be more important than ever, and now it Avaya doesn’t have to look outside its Emerging Technologies group for what it considers to be the best-in-class technology. At the same time, the company brings a robust set of analytic tools into its mainstream product line and will add an important component to the quality of service that it delivers to its enterprise customers and, in turn, to their customers.

The highly visible mandate for companies to do a better job of listening and responding to their customers and prospects have attracted a number of non-traditional competitors to the monitoring and analytics market, most notably Cisco Systems, with its “Socialminer” product has elevated the visibility of some specialty firms like Nexidia, Calabrio, Utopy and a handful of others addressing the challenge of capturing and analyzing input from a number of sources (IM, Chat, Text, Web, phone), encompassing a number of modalities (speech and text).

Comverse’s sale of Verint is complicated by a number of factors. Verint already trades on the NASDAQ and has a market cap of $1.3 billion. Before Comverse can divest its share of the company, it has to get up-to-date on long-delayed SEC filings – a process that is expected to be completed in March. That should be good timing because interest in faster, more comprehensive capture and analysis of user generated metadata is building.

Nexidia Capture, by contrast, supports “complete analysis of recorded interactions.” In conjunction with Nexidia’s ESI (Enterprise Speech Intelligence) product suite, companies can conduct the sort of mining and extraction of meaningful business intelligence from multiple audio streams. This could also support a service like Cisco Pulse which can do pattern matching among multiple audio streams to detect commonalities and help build communities.

Nexidia will be demonstrating its integrations at this week’s Cisco Collaboration Partners Summit in Phoenix. I will be there and will be getting some real time demos.

Instead of “speech” the center of gravity has decidedly moved toward creating the optimal multimodal customer experience. Thus the idea of Avaya re-naming its long-standing “Voice Portal” the “Experience Portal” or Microsoft launching a company-wide marketing, development and packaging initiative around the idea of a “Natural User Interface” that includes speech but adds typing, texting, and “gestures” show that speech has a gigantic future – but it is as part of a combination of technologies brought to bear to solve the needs of end-users.

Past SpeechTEK’s were specific to speech. When we talked about standards, it was largely to make ASR recognizers and TTS synthesizers work and play well with one another. Likewise, analytics were “speech analytics” put into play to find the faults in spoken dialogues or root cause for abandoning an endless session inside an IVR. ADEs (Application Development Environments) governed “dialogue modules” or reusable code to create better speech-based interactions. You get the idea. Now, in the spirit of RC (Recombinant Communications) developers are retaining the best of their speech-based experience while leveraging their investment in the logic and workflow encased in their Web sites and mobile apps.

SpeechTEK made it evident that, while speech processing remains an important and necessary resource for today’s self-service and assisted service applications, we’re well beyond such a one-dimensional treatment of reality. User Experience is the rightful center of attention and the speech community has come around to recognize that it cannot address the full spectrum of user options by itself. That’s why most of the briefings we took at SpeechTEK addressed how vendors are addressing the demand for “cross-channel” and “multimodal” communications.

More details in next week’s advisory on “Automated Speech in the Multimodal World”. So many opportunities… So much time!

Non-Clients – Click Here to View the Report Summary