Sponsored By

Siri Meets WatsonSiri Meets Watson

A hosted solution leverages conversational virtual agents powered by artificial intelligence to automate a variety of inbound and outbound call types.

Sheila McGee-Smith

March 22, 2012

3 Min Read
No Jitter logo in a gray background | No Jitter

A hosted solution leverages conversational virtual agents powered by artificial intelligence to automate a variety of inbound and outbound call types.

VPI announced today what they describe in the press release as “game-changing” virtual call agent technology. VirtualSource is a pay-as-you-go, hosted solution that leverages conversational virtual agents powered by artificial intelligence to automate a variety of inbound and outbound call types. VPI's website lists about 30 different interaction types that are already available, from taking orders to processing returns and customer surveys. I was intrigued enough listening to the demo calls to request a briefing.

I doff my hat to VPI's VP-Marketing Patrick Botz. In the course of a 30 minute call, he attached this announcement to just about every hot button in the contact center and customer experience space today…and then some.

Cloud: VPI has launched VirtualSource as a hosted solution. Similar to other cloud-based solutions, this shortens deployment time and provides companies with an OPEX versus CAPEX solution. Companies can deploy VirtualSource in a hybrid model, allowing the existing premises-based IVR to handle some interactions while specifying those that are sent to VirtualSource.

Virtual: Admittedly, this is not the trendy "virtual" of data center deployment of servers or desktops. Virtual here refers to an automated agent persona, part of the reason for the comparison in the title to Siri and Watson. Asked how VirtualSource is different from virtual agent solutions that have been around for years and not really succeeded, Botz highlights the embedded artificial intelligence that is layered on top of the Nuance-based speech technology. Instead of using decision trees, which must have rules and branching created for every possible response, VirtualSource learns on its own.

Mobile: Mobile is hot, with companies spinning like tops trying to better meet the needs of their mobile customers. Just like Siri can simplify the execution of simple tasks for a consumer with a smart phone, Botz contends that VirtualSource can simplify the completion of routine tasks for businesses. Without requiring a special app be downloaded from an app store, companies can improve service to customers on mobile devices.

Minimal Professional Services: One of the problems with sophisticated solutions is the time and expense of the professional services that must be applied to get an acceptable result. Think CTI and its lack of market traction in the 1990s. When I asked about professional services costs, I was completely taken aback by the response. Initially, there are no professional services charges. Customers pay 25 cents per minute and sign a one year contract with a monthly minimum. Because VirtualSource learns as it goes instead of requiring that rules be written, Botz says that with as few as 10-20 sample calls and the scripts used to train agents, an application can be up and running.

An example reinforces the no-programming-required nature of VirtualSource. The link below plays a demo of a call to make payment arrangements. You’ll hear the caller negotiate with the virtual agent on when the second of two payments will be made. When I asked Botz how the virtual agent is empowered to negotiate, he replied that VirtualSource learns from the sample calls or training scripts what the acceptable leeway is and applies it to each call. Fine-tuning is performed by VirtualSource for typically 4-8 weeks--again with no fee.

The only thing missing was a tie-in to social media--but I'm sure Patrick will start working on it as soon as he reads this. For more on the latest in social, cloud and mobile for the contact center, catch one of the three sessions I'll be running next week at Enterprise Connect.

About the Author

Sheila McGee-Smith

Sheila McGee-Smith, who founded McGee-Smith Analytics in 2001, is a leading communications industry analyst and strategic consultant focused on the contact center and enterprise communications markets. She has a proven track record of accomplishment in new product development, competitive assessment, market research, and sales strategies for communications solutions and services.

McGee-Smith Analytics works with companies ranging in size from the Fortune 100 to start-ups, examining the competitive environment for communications products and services. Sheila's expertise includes product assessment, sales force training, and content creation for white papers, eBooks, and webinars. Her professional accomplishments include authoring multi-client market research studies in the areas of contact centers, enterprise telephony, data networking, and the wireless market. She is a frequent speaker at industry conferences, user group and sales meetings, as well as an oft-quoted authority on news and trends in the communications market.

Sheila has spent 30 years in the communications industry, including 12 years as an industry analyst with The Pelorus Group. Early in her career, she held sales management, market research and product management positions at AT&T, Timeplex, and Dun & Bradstreet. Sheila serves as the Contact Center Track Chair for Enterprise Connect.