I've been a healthy skeptic of UC for a number of years, having been prejudiced by the Nortel/Microsoft effort to promote the concept and then watched as numerous IP PBX and "cloud" players evolved themselves into UC companies. The truth is more complex, since voice continues to be a silo for the majority of the market.

Consider any typical UC offering being promoted today.  You get email, IM, presence, telephony, messaging, video as the core applications, with voice sitting neatly in its own little package with voice mail and visual voice mail (speech-to-text via email) complementing stock phone functions such as call forwarding, caller ID, and simultaneous ring.  I'd be more impressed if you got an IM with caller ID information cross referenced with any address book information when you had inbound calls set to "off," but so-called state of the art is getting a voice mail/visual voice mail notification with just the inbound phone number and time stamp on the subject header.

The level of processing intelligence used on any sort of inbound voice – either live or recorded – just isn't there.   Tools such as speech-to-text, Hypervoice and voice analytics are not common, with most uses limited to large call centers. 

Call recording and voice mail are two separate applications – why? Both applications record voice, with the latter just having one party on the line.  Shouldn't you have one "unified" call recording application?

Cloud services and the migration to an all-IP voice should make for a more unified communications solution, but the voice-based world needs to pull out of its collective "Voice is Dead" funk and start aggressively rethinking how to process voice as data, rather than sticking with a broken analog telecom mindset that continues to silo voice.

What the world ultimately needs is a "Siri with rules," a call agent that can be set to screen inbound calls using a mixture of caller ID, contact/address book information , and the ability to process through what is being said and take actionable steps.

For example, if someone from TMC calls me to talk about the Smart Voice conference, I would want to get some sort of "immediate" alert if the voice message used a key phrase like "There's a problem" or "There's a big problem."  If we wanted to get fancy, the level of immediacy would go up depending on other key words such as "immediately" and "as soon as possible."   Analysis of the voice stream for emotional tone -- as currently done by large contact centers via voice analytics -- would also set the immediate level for the appropriate IM.

The trick to layering more intelligence onto voice calls is making it what I'd call "Apple simple." You want to get enough functionality to have people easily use most functions, with a back-end API available to develop things like the abovementioned immediacy level to send alerts.  

If you want a big picture view, just think about how many cloud communications UC services business are on the market today.  Cloud voice is a commodity.  To keep on growing market share, differentiating services, and adding revenue per user, service providers are going to have to add Smart Voice services or try to compete on volume and price.

If you're interested in the future, I'd like to ask you to come to the Smart Voice Conference on January 28, 2014, at IT EXPO East in Miami, Florida.  We've already got a number of industry experts in call recording, cloud voice services, voice biometrics, voice analytics and Hypervoice speaking with Martin Geddes, co-founder and executive director of the Hypervoice consortium, delivering the kick-off keynote.