Q&A: David Morand on Integrating a Contextual AI Assistant with VoiceXML
Amazon and Google may have made it easier to develop speech applications without VoiceXML, but the technology is still widespread in the industry, especially in large contact centers. If you need a dialogue engine that will allow you to develop next-gen conversational IVR applications while being compatible with the VoiceXML standard, you need to hear what David Morand, Senior Software Developer, Nu Echo, has to say.
Q: What is the essence of your presentation, "Integrating a contextual AI assistant with VoiceXML," at the SpeechTEK Conference, April 27-29, in Washington, D.C?
A: We have explored dialogue management capabilities of the open-source AI assistant solution offered by Rasa and its integration with the VoiceXML standard. We will describe the rationale behind the Rasa-VoiceXML solution, the use cases that were explored, and how we leveraged both deterministic dialogue strategies and machine learning capabilities. We will also cover how we integrated its open-source components with VoiceXML, including multilingualism, audio file management and speech grammars. Deployment and testing strategies will also be presented.
Q: What is the context and why is it important?
A: While many options are available to develop speech applications without VoiceXML (Amazon Lex and Google Dialogflow come to mind), it is still widely used in the industry, especially in the large contact centers which are our primary customers. We needed a dialogue engine that would allow us to develop next-gen conversational IVR applications while being compatible with the VoiceXML standard.
Q: VoiceXML has been widely used for many years. What did you feel were the shortcomings of VoiceXML?
A: VoiceXML biggest shortcoming is the lack of control structures which make it unsuitable as a single tool for developing complex IVR applications. That’s why for a long time we used VoiceXML as an intermediate language to access ASR, TTS, telephony and media resources on top of a Java application development framework (Rivr) which was open-sourced several years ago.
Q: Why did you select Rasa for integration with VoiceXML?
A: Rasa offers us a great mix of ready to use tools, dialogue strategies (ML and deterministic) and deployment recipes to quickly create working intelligent assistants. Most importantly, it is fully extensible and customizable which allowed our engineering team to adapt it to support VoiceXML and the specificities of speech applications using Python.
Q: What are the challenges of integrating Rasa with VoiceXML?
A: Incorporating foreign VoiceXML concepts in Rasa required some outside-the-box thinking. Also, properly managing multilingualism and dynamic audio concatenation was a bit tricky. For the details, you will have to come to the presentation.To see presentations by David Morand and other speech technology experts, register to attend the SpeechTEK Conference.
Learn how conversational technologies are changing the way we bank, and how Discover is embracing this trend to better support companies and employees.
Current AI and machine learning (ML) technologies are starting to change the way we build and innovate. However, the power of our current ML technologies is not fixed. Sam Ringer will explore where ML is at the moment.
The fact is, many consumers, and in particular younger ones such as Millennials and Generation Z, prefer voice communications when interacting with organizations for urgent support.
Businesses let hundreds of millions of dollars slip through their fingertips because they fail to recognize opportunities embedded in conversational nuance. But AI can help.
Allstate Conversational Designer Katie Lower outlines working models for assessing the viability of a conversational interface with multiple teams within an organization in this clip from her presentation at SpeechTEK 2019.