I recently visited the SpeechTek 2010 show in New York, where I met with a number of vendors focused on improving state-of-the art technology. Most memorable, by far, was Microsoft (News - Alert), because they showed me deep speech integration into the operating systems of both the Windows 7 phone and Xbox. In addition, they showed me Xbox Kinect, which is a $150 device that plugs into the gaming console and, once connected, allows your body gestures to control games. The system uses camera and is very accurate. I would suggest the word “revolutionary” was coined to describe what it does to computer interfaces.

Getting back to speech, Jeff Schlueter (News - Alert) vice president of marketing and business development, at speech analytics provider Nexidia explained how contact centers consist of more than 50 percent of the company’s growth, and banking, cable and healthcare companies, along with government regulators, are proving to be strong customers as well.

One way the company helps call centers is with its language assessor prescreening tool, which evaluates new call center hires via phone before they are hired. The company recently renamed and repackaged its products into the following buckets: Discovery Edition, Analysis Edition, Quality Management Edition and the Enterprise Management Edition, with the goal being to more seamlessly connect customers with appropriate solutions, based on their size and requirements.

Moreover, the company is working on productizing multichannel analytics, which will allow companies to better determine what customers are thinking by trolling their Web and speech interactions. The company raised $1.5 million in financing recently and no doubt some of the proceeds are being used to further this cause.

Schlueter has interesting insight on why speech is a better gauge than text for gauging customer perceptions. He believes it is easy for someone to flame a company quickly online, but says it is pretty easy to analyze such content, since it is text-based. Speech, on the other hand, is more accurate, he believes, as people tell you more about their true feelings in a conversation. The challenge, of course, is deciphering content. That is where their company comes in, and we can look forward to a new product from them at some point in the future.

You really can’t mention speech technology without mentioning Nuance. The trio of Dena Skrbina, Laura Marino (News - Alert) and Andrea Mocherman told me how the company has evolved from a speech company to a full provider of customer care, including cloud-based solutions. Nuance (News - Alert) is working on improving customer interaction with automated systems and, to that end, they are happy to be improving the comfort level of speech interaction with their recently released Dragon 11 Dictation Engine, they say is the most accurate product in the series.

An interesting part of the dialogue was that comfort with speech will soon approach our comfort with search engines like Google (News - Alert). The idea is that, when users are confronted with speech systems, they typically utter one-word answers (as they have been trained to do). Similarly, users were once used to putting single words into search engines but, over time, became accustomed to long tail queries.

Nuance believes the world of speech callers will become more comfortable uttering phrases like, “Why is my bill so much larger than last month?”

Another area of focus for the company is outbound notifications, where it is powering services for Canada’s ScotiaBank, allowing customers to set InforAlerts when their bank accounts hit certain thresholds. In all, there are 15 possible alerts (security alerts are free) and 90 percent of customers are using these services and paying for them. Perhaps one of the reasons Canadian banks made it through the housing crisis unscathed was they were focusing on tech solutions to help customers and boost revenue, instead of packaging and repackaging soon-to-be-worthless documents that represented risky mortgage bets.

Another area of focus for the company is the smart grid, where the company is working with utilities to notify customers of lower utility rates

But, perhaps most interesting is the company’s integration into smartphones, where it is working with carriers to place apps on phones that are activated certain calls are placed. For example, when you call your mobile provider, a menu pops on your cell phone allowing you to quickly navigate to a page where you can add a second line or pay your bill.

We certainly aren’t at the point yet where computer speech recognition is going to pass for human interaction, but we are getting close. Moreover, as technology continues to improve, it allows more innovative services and better service levels, and lets regulators to more accurately monitor companies’ activities.

I’d say the speech technology market is fairly healthy at this point and we are certainly at a point in time where analytics is good enough that companies can really mine their recorded and live phone archives to factor into their BI strategies. Moreover, corporations will save money as they shift more calls from live agents to automated systems.