By Steve Liang, AVA Project Lead
Today we’re excited to introduce a beta of AVA, VMware’s Automated Virtual Assistant. We designed Ava to help users of VMware Docs find answers to common product questions. Using natural language processing, Ava is trained to understand key VMware concepts and recommend further reading that will answer your questions. Ava’s recommendations come from a combination of product documentation and FAQs, with more content being added all the time.
Using Automated Virtual Assistant
Because Ava is still learning, we’re only inviting My VMware users to interact with Ava. So to meet Ava, you’ll need to log in to VMware Docs using your My VMware account and launch the chat icon from any page. You’ll quickly notice Ava is in “Training Mode.” Right now, your interactions help Ava learn more about frequent issues and questions. Training Mode allows Ava to start associating the many variations of the same questions and search terms to similar recommendations. You can ask Ava about using VMware products or describe problems you might be having. Ava might ask clarifying questions before returning the best answers to resolve your query.
With Ava’s in-context assistance, it’s easy to toggle between a deep dive on answers and continuing the conversation. You can continue to browse content on VMware Docs and refer back to Ava at any time to pick up where you left off. Additionally, to provide a consistent experience, Ava retains conversation context, so if you ask, “How do I tune the performance of a VCPU in vSphere 6.5?”, Ava will remember the product and version for follow-up questions. You can reset the context to ask new questions about a different product any time you want.
Ava’s goal is to help you resolve problems as quickly as possible. Recommendations are driven by a combination of semantic similarity and feedback from you. We added a built-in feedback mechanism so you can contribute to Ava’s training. Every time Ava gives you a set of recommendations, you can tell us which one was the best answer. This helps Ava improve the support experience and gives us valuable information to tune Ava’s continued learning.
At VMware, we’re committed to making your product experience valuable and engaging. Over the coming months, we’ll continue to deliver frequent updates to Ava to expand the features and functionality. With this beta launch, Ava is available in English, but we are working to support additional languages. Your feedback is key to continue improving Ava, so let us know your thoughts through the Ava feedback.