Q&A: Walter Rolandi Talks Dialog Development Tools
Times, they are a-changing—and nowhere is that truer than in any tech space. Even developers can get whiplash trying to keep up with the latest and greatest advances in their particular niches. With that in mind, Speech Technology magazine interviewed Walter Rolandi, Ph.D. of The Voice User Interface Company, to talk about new dialog development tools versus traditional speech IVR technologies and how they measure up.
Q: Can you tell us a bit about traditional speech IVR technologies?
A: Traditional ASR applications employ specific grammars in every interactive state. The recognizer can recognize only what the VUI designer includes in the grammar. Any other user utterance made in the state will generate an error. In this sense, traditional ASR applications can be fairly brittle.
Q: How are dialog development tools changing and why?
A: The traditional way of creating a dialog, especially the traditional way of specifying what is to be recognized is undergoing a fundamental change.
Q: What are the primary differences between the “old” and “new” tools?
A: Older tools permit designers to specify and test specific entries in a grammar. More modern tools allow designers to provide examples of the sorts of things that should be recognized in a particular application. The examples are associated with a particular “intent.” Using artificial intelligence technologies, modern tools can generalize the examples and allow thousands of alternative ways to recognize that same intent.
Q: What do you see as the advantages of newer dialog development tools?
A: The single greatest advantage of the new approach is its tolerance of vastly variable ways of saying things. This can make a dialog not only sound more natural but it can also dramatically reduce the incidence of setbacks in the user experience due to speech recognition failures.
Q: Are there any disadvantages?
A: There is definitely a learning curve. Instead of specifically defining exactly what will be recognized, the VUI designer must learn to “train” the application to “fire” when a particular “intent” is evoked.
Q: When a company is looking for a new tool to work with, what should they consider before making a commitment?
A: They should ensure that they can approach the task with an open mind and willingness to adopt a different development paradigm.
Poorly designed IVRS have been angering customers for decades, but it doesn't have to be this way. We talked to the Founder and CEO of UJET about how well designed IVRs can improve customer experience.
Speech applications are hot, but still provide more value than revenue. We talk to the experts about how to monetize your speech apps, and how value might be just as important as revenue.
As speech technology becomes more commonplace, new markets are opening up for vendors. But as new opportunities abound, how do voice vendors reach out to potential new customers like marketers?
Just as brands are perfecting DX, they have to rethink every aspect from search to delivery. Eli Finkelshteyn, CEO and co-founder of Constructor.io, weighs in on the challenges brands face in the voice arena.
At the 2019 SpeechTEK conference Yves Normandin of Nu Echo, Inc. and Deborah Dahl of Conversational Technologies, will present "A Comprehensive Guide to Technologies for Conversational Systems." Conference chair Jim Larson talked to Normandin and Dahl to get a sneak peek of the session, and learn about conversational system technologies.