Best practices for a successful virtual assistant deployment

Learn how to develop AI-based VAs to minimize risk and reach the technology’s full potential

Each conversational AI project has technical, timing, and economic risks, largely due to the high expectations of virtual assistants. Traditional approaches to software development need to be adapted to alleviate the risk. To solve this problem, we introduce the S.A.F.E. method as standard guidance for developing AI-based VAs.

In this video, you’ll learn the 4 basic capabilities your virtual assistant needs in order to be successful and deliver a solid ROI:

• Smart: Learn about Level 3 AI to give virtual assistants the ability to manage non-linear, context-sensitive, multi-round conversations.
• Action-oriented: How can virtual assistants tackle entire workflows, not just simple Q&A?
• Fast learner: How do you create a virtual assistant who can easily ingest corporate knowledge, learn new skills, as well as get on the job training.
• Extensible: What type of platform do you need to support any use case, any channel, as well as encourage collaboration with human agents?

Watch Prof. Yi Zhang, CTO of Rulai and professor at UC Santa Cruz, discuss how to go about developing powerful virtual assistants that deliver a solid ROI. Prof. Zhang has over 20 years of experience in AI, has been a consultant or technical adviser for several large companies, and received various academic rewards for her work.

This 30-minute presentation was adapted from a speech given at the 2020 Chatbot Conference in New York.