The Reasoning Show

The Reasoning Show - Reasoning through the AI Revolution

The industry's leading AI podcast, explores how leaders think through AI, technology, and transformation. Each week founders, investors, and operators unpack the decisions behind the systems shaping modern business.

Hosts: Aaron Delp and Brian Gracely

All Episodes

The Reasoning Show

Natural Language Understanding with AI for IT Support

July 15, 2020 • Cloudcast Media

0:00 | 27:59

Vaibhav Nivargi (CTO & Founder @Moveworks) talks about Natural Language Understanding (NLU), interacting with users using chatbots, and augmenting customer service with AI.

SHOW: 458

SHOW SPONSOR LINKS:

Datadog Security Monitoring Homepage - Modern Monitoring and Analytics
Try Datadog yourself by starting a free, 14-day trial today. Listeners of this podcast will also receive a free Datadog T-shirt
strongDM Homepage
Start your free 14 day trial today at: strongdm.com/cloudcast

CLOUD NEWS OF THE WEEK - http://bit.ly/cloudcast-cnotw

PodCTL Podcast is Back (Enterprise Kubernetes) - http://podctl.com

SHOW NOTES:

Moveworks Homepage

Topic 1 - Vaibhav, welcome to the show. Tell everyone a little about yourself and what got you started in the AI space?

Topic 2 - We’ve done a number of AI/ML shows over the years, but we haven’t talked much about Natural Language Understanding or NLU. Let’s start there, can you give everyone an introduction?

Topic 3 - Based on that, is the primary interaction with the end user through a chatbot or something similar? What are the primary use cases and tools you are seeing in the industry? Is this a Slack and/or Microsoft Teams integration? Unsolicited plug, I’m a customer in my day job…

Topic 4 - We’ve been talking a lot on the show recently about the migration to SaaS based products. What is the model here? Is the AI central (cloud hosted) or private and in-house? Do you have the concept of a template AI and then each customer AI is an instance or is this a central AI that is called? How does it get customized and updated over time? What training is typically required and is this training on-going?

Topic 5 - How do you prevent user frustration from “loops” or unanswered questions? I think of the voice automated telephone systems I’m not a fan of as an example. How would you handle language that isn’t built into the AI?

FEEDBACK?