The consensus opinion is that Amazon will be launching its own “Siri Killer.” But that doesn’t fully explain the level of secrecy and subterfuge surrounding the initial acquisition. The dime dropped on Yap on September 8. A few weeks later, those of us who were participating in the beta of Yap’s Voicemail-to-text app (on the iPhone) were informed that the service would be discontinued as of October 20. Then there was something of a radio silence from the folks in Charlotte.

I don’t like to think of any product or service as a “killer.” When we look back on the 4th quarter of 2011, it will be very clear that, whether it succeeds, fails or just goes off line periodically, there is no question that the introduction of Siri on the iPhone 4S irreversibly changed the mobile device and services landscape. In preparation for the holiday buying season, the general public should be thrilled to find a wide variety of smartphones and tablets for sale, each with a price range, feature set and user interface designed to attract new customers – the non-geeks.

In this respect, the ability to understand spoken instructions or dictated messages “reliably-enough” at home, in the office, on noisy streets or public venues like airports is turning out to be table stakes for achieving mass market potential. We’re moving into the Star Trek-like world of talking tablets and. to avoid dependencies on third-party technology providers or “speech specialists” is now part of the Prime Directive.

There is no mention of the Yap acquisition on Amazon.com’s Web site, or in its most recent 10-Q filing at the SEC (dated October 26), nor is there a mention of an acquisition in other SEC filings. Therefore, the price paid must not have been material to Amazon.com’s general financial situation, the transaction may have been largely non-cash in nature, exchange of services or what amounts to a refinancing.

Much is made Yap’s proprietary technology. It’s “Speech Cloud” technology promised highly accurate rendering or transcribing of spoken words. It was a fully-automated solution, like Google Voice, but apparently was achieving high levels of accuracy without the benefit of Google-like scale for its corpus of captured utterances. Opus Research believes that its core speech processing engine is the AT&T Watson Speech Engine, though it hasn’t made that formal announcement and the Yap’s co-founder Igor Jablokov was a veteran of the Pervasive Computing division of IBM, where much of the speech processing technology now licensed to Nuance was developed.

Holiday shoppers will be presented with a wide variety of options and prices for devices that act as mobile assistants. Siri will be running on all manner of iOS-based devices. Nuance is offering Dragon Go! on iOS and Android devices as a very results oriented mobile assistant. Vlingo has been expanding its service offerings and geographic coverage. And now we can look forward to an Amazon.com branded product – most likely part of the Kindle family – that will leverage the Amazon.com e-commerce infrastructure (marketplace and checkout), understand commands and dictation, and probably offer more options for audio books.

Now our Prime Directive is to get more comfortable with all those Speechable Moments to be had.

Syclo, founded in 1995, has a history of working with SAP, IBM and other major enterprise infrastructure providers to extend corporate network services to mobile workers and customers. Its flagship “platform” called Agentry is now integrated with West Interactive’s customer care platform in order to provide West’s customers with ways to interact with mobile customers that has proven to be reliable, scalable and secure.

“Mobile fragmentation” (resulting from the existence of multiple operating systems on smartphones, as well as the staying power of feature phones) has been the bane of the application development community. Over the years Syclo’s has created a software development tool that co-founder Dave Kleban told us enables programmers to create scripts or programs that run as “native apps” without writing any lines of code. Support for multi-modal customer care strategies is a clear benefit.

West Interactive’s programmers can “orchestrate” the end-to-end customer experience across many touchpoints. For example, in the utilities vertical where they have many clients in common, the orchestration of a service call (truck roll) can be accomplished more smoothly when scheduling, notification, confirmation or facilitating changes can be carried out using the medium of the customer’s choice.

The Syclo partnership comes on the heals of a deal with Radian6 (now part of Salesforce.com) to define a way for West’s clients to incorporate feedback from social networks into the customer care loop. It shows that West is willing to identify and forge relationships with “best-of-breed” technology providers to bring its clients into the age of social/mobile aka Conversational Commerce.

Apparently, in the case of freely provided API access, popularity and frequent amount to “abuse” and certainly lead to an economic burden of sorts. In this post, Adam DuVander at ProgrammableWeb notes that his company’s directory of mashups include 34 services that employ the Google Translate API. This includes “Chat Translator and Speaker for Skype,” which translates instant messages in real-time and uses text-to-speech to “read” incoming messages. I also observed several mobile “speech-to-speech translation” services using Nuance Dragon Dictation on to front-end Google Translation services and then render the output using text-to-speech. My impression was that Google already put a limit on the number of translations it would execute for individual service providers.

I’m having a deja vu about the unilateral discontinuation of a popular service by Google. It is very similar to GOOG411, a freely provided dial-up directory assistance service, which Google stopped offering in November. At the time, I asserted that the service had provided Google with what it needed (a collection of utterances to help fine tune its speech recognition for other products like Voice Search and Voice Actions. At the same time, it was an acknowledgement that it no longer would support free call completion (and phone calls) by the general public. Indeed, the company has been waffling about charging for calls originated through Gmail as part of the Google Voice feature.

Unlike GOOG411, 3rd parties were building their own services and business plans around the Translate API. For Google to discontinue the service unilaterally will either cripple their business plans or send them to another translation API provider. ProgrammableWeb lists a number of automated and human-assisted translation APIs here. Other than Google (which accounts for about six flavors of translation APIs on the list) Microsoft/Bing being the most recognizable brand.

In this blog post, Dana Blankenhorn urges Google to come clean with a longer explanation. Was the service withdrawn do to government pressure? Does it signal erosion in Google’s cost advantage over competitors? Will Google flat out discontinue translation services?

As an alternative to deprecation, I can easily imagine Google re-introducing the service as a product that carries a fee. Today, the API is still part of Google Labs and not subject to charges. My guess is that they will harden it in some way, attach a meter to count the “dips” into its database and transform those “abusers” into customers by attaching fees per word, translation or bundles of translations.

In over 80% of instances, mobile users download a female voice for their device or application. For English speakers, the figure is closer to 85%. By the way, the combination of US and UK English accounts for more than half of all downloads, with a cluster of Russian, French and German accounting for something on the order of 6% each.

The predominance of English-speaking applications reflects the relative distribution of smartphones in general (and Android in particular) as well as the mix of applications that are speech-enabled. The preference for female voices is a bit of a reversal for the speech processing industry. Mobile ASR (automated speech recognition) is notoriously biased against female voices. At first it was written off as the product of grammar development built primarily on male utterances. Later there was a mismatch between the range and pitch of female voices and the sweet spot in the phone lines’ (and wireless links’) information carrying capabilities.

This gender gap may be short-lived or, ultimately, of little meaning. Choice of preferred voice is either application driven (most navigational applications and devices use female voices) or developer driven – meaning it is out of the hands of the end-user. The next generation of “life-like” TTS is designed to be extremely flexible to the point where “publishers” or entertainment content providers can assign specific voices to individual characters. In these cases, the mix of gender will move to reflect the general population – something close to 50/50.

Today Nuance announced that its developer support program is expanding to China by adding Mandarin Chinese to a set of languages that already supports US and UK English, European Spanish, European French, German, Italian and Japanese for dictation and search, as well as 35 languages for text-to-speech conversion.

While mobile and multimodal Web application platforms (like Appcelerator, AT&T and others ) have attracted in excess of 2 million developers, it has proven much more difficult for mobile speech processing specialists to achieve such lofty levels of registration, but Nuance can claim showcase apps that include the AI-based mobile assistant Siri, Price Check by Amazon, Ask for iPhone, as well as other vital reference and memory aids from Merriam-Webster, Dictionary.com, SpeechTrans, Avantar (AirYell), Sonico (iTranslate), and Catalyst (Taskmind).

Opus Research strongly believes that support of the developer community is key to broadening the appeal of speech-enabled mobile applications. As we’ve noted, the 2,500 folks who registered to use Nuance’s developer resources is a small fraction of millions of folks who who hope to make their mark as mobile developers in the white hot world of mobile gaming, socializing, search, check-ins and e-commerce. Our belief is that it won’t take long for the community’s creativity to turn to multimodal input, and that bodes well for both Nuance, Google and any other platform that showcases accurate speech recognition and bearable text-to-speech rendering as part of the user experience.

Nuance’s formal entry into China expands mobile speech’s global footprint.

TomTom is, by no means, the only horse in the natural language race. Nuance Communications, for instance, has VoCon “One Shot Destination Entry” baked into the nav systems in prestige nameplates, including Audi, BMW, Ford and Mercedes-Benz. In February, the company began offering a pre-packaged SDK (Software Development Kit) called VoCon Navigation, which it calls a “ready-to-integrate pre-packaged voice application module for embedded one-shot destination entry.”

As we see state troopers and local law enforcement cracking down on distracted drivers, the simplification of destination entry for navigation systems is becoming a valuable attribute.

The beginning of the video is a fairly standard demo of the Vlingo Virtual Assistant as invoked by “double tapping” the home key [although the gardening gloves provide an interesting twist that would preclude the use of fingerprints for user authentication]. Forty-four seconds into the presentation, attention turns to the in-car experience with a demonstration of “truly hands-free” service initiation using the trigger phrase “hey Galaxy!”

The demo worked in what looks like a crowded auditorium, which is always a big plus. And the spokesperson was able to demonstrate how well a wireless device can be activated in response to a trigger phrase, “understand” and parse commands from content (so that “Text to Andy” is not transcribed but “Meet you in the office at 3 PM” is). To my mind, what’s left out is the obvious next step of the system asking “Would you like me to add this to your calendar?” which always makes for another aha! (speechable moment) for audience members who have not been exposed to the mobile apps that serve as personal organizers and memory aids.

Today, Vlingo counts Google, YellowPages.com, foursquare, Fandango, Kayak and OpenTable as partners in its efforts both to complete tasks and to monetize usage. By the end of the year, the roadmap will lead it to form more partnerships in the European theater. In the spirit of multimodality and multi-device support, the company also spelled out plans to grow its presence on tablet devices beyond the Samsung Galaxy and to grow the base of applications and services that drivers take advantage of in their cars. The company expects to fulfill the promise of a “voice remote controls” in conjunction with major TV manufacturers and its product managers expect to see the venerable PC playing an important role by providing the same Virtual Assistant services that are available on mobile phones and in cars.

Vlingo also plans to expand geographically by adding languages to its roster. It already supports English, French, Italian, German, Spanish and Mandarin but has plans to add more (unspecified) languages in calendar year 2011.

Waze was one of the earliest mashups of crowdsourcing, Maps and Navigation services on an application for Apple’s iOS based devices. It has, over time, added versions for Symbian, Android, Blackberry and Windows Mobile based devices.

As noted in the video above, its designers have long contemplated adding more “voice” to the app, and they have been true to their word. It’s clearly not the dream “hands-free, eyes-forward” use case that would make it most appropriate for drivers. But it is very cool to have audio “alerts” that are the product of spoken input from drivers providing contemporaneous commentary on the traffic or weather phenomena that they are observing. The company has said that upgrades to the other platforms will be available in the future.

Both input modalities provide quick, convenient ways for holiday shoppers to comparison shop, spanning the World Wide Web and the Real World. MeMeMe, as the verbiage on its Web site indicates, is on a mission to “voice-enable any mobile application or networked service.” That makes it direct competition to Nuance, Microsoft/Tellme, AT&T/Watson, Loquendo and a few others. The Buy.com app for the iPhone is a good demonstration. It also shows how speech-based input is a peer to (but not a replacement for) keying in info or scanning barcodes.