Watch videos with subtitles in your language, upload your videos, create your own subtitles! Click here to learn more on "how to Dotsub"

10893_AI_Business_School_Lili_Cheng_FINAL_V3

0 (0 Likes / 0 Dislikes)
[AI Business School] >> Lili: Today I'm going to talk to you about conversational AI, one of my favorite topics. It is one of my favorite topics because we believe that what people say, type, and the way they communicate with one another and with their services impacts the way everything will be built in the future. So everything you already type often shows up in your keyboard on your phone, and the way you interact with customers, the way you work inside your company, this is all going to get better as software can do more and understand more about the meaning of the words that you say and do. About two years ago, we began the bot framework, and we have seen incredible growth in this space. Using bots and conversational apps today, we have over 500,000 developers using our open-source tools, and over 10,000 companies building conversational AI experiences, either for their customers or to help their employees and partners get work done better together. So that's super exciting. We've really loved working with UPS, and one of the great things about working with UPS is they were able to get a solution up and running really quickly. So it's hard to know how to get started with AI. And starting with a really concrete problem that you might have or an opportunity to interact better again with your customers is a really simple way to get up and running. And here UPS was able to get up and running in just a few weeks, roll this out to their customers and really begin to learn how this experience helped people get their jobs done more effectively. Just to back up a little bit, what are the conversational AI principles that are really guiding how we prioritize the work and our products? First, we want you to have complete control over the user experience, and really own your AI future. We want you to have a lot of flexibility in the way that you integrate IA tools to do conversation, and for you to really own the brand. The AI story for your company is so much about who your company is and how you converse inside and outside. So you really need to own that user experience and you need to have complete ownership and control of the data. And you want to be able to make sure that any conversational experience you build works everywhere you interact with your customers so you don't have to keep re-tooling and rebuilding as you have a new experience. If you have a new app, if you have a new web page or if your customers are interacting with you in many social experiences, like Facebook or Slack or in teams or in email. We have the richest, most comprehensive tooling set across the life cycle of your project, whether you're planning, you're building it, you're coding it, you're testing it, you're publishing it, you're connecting it to sources, or every day as people use it, you're evaluating usage and you're making it better. Our tooling is very, very rich, and today we'll only touch on a few pieces of this tooling but we're really excited because this is also extensible and we've been able to work with many other solution providers to continually make this tooling richer. One of the core pieces, probably the core of our conversational AI for people building experiences, is the Azure Bot Service. The Azure Bot Service leverages the web app that many people are already using to build apps, and really makes it easy to language-enable a real-time conversational experience hosted on Azure. So if you want to build a conversational experience, this is a great place and one of the key elements that you should really try out. Our Azure Bot Service. We integrate really well with all of our cognitive services. So Vision, which includes taking pictures of text and understanding the text in your images, as well as face, and lots of other cool stuff with Vision. Speech is obviously really important. All the language understanding, including translation and real-time translation, as well as really rich knowledge that we have had so much experience working with Bing and understanding the world's information. If you don't want to code, we make it really easy for you to get going with the QnA Maker service, so if you already have a FAQ or an Excel spreadsheet with questions and answers or a document, you can really easily just point to that web page or document or upload it and author conversational question and answer bot just in minutes, leveraging all the data that you already have. So QnA Maker is a really neat cognitive service. Embedded in the QnA Maker is some really interesting AI. So we see that a lot of customers have documents or manuals, in some cases, over 20,000 manuals, and they're often PDFs or in a variety of formats with very complex information, like tables, that they really want to be able to generate bots over, conversational experiences over. So we have really rich, really neat, leading edge AI tooling embedded in a very simple and easy-to-use tool like QnA Maker. So this is just an example of the kind of tables that we can generate question-and-answer pairs over. Obviously, speech and text are great, but we also have video. So we can start to do things in Video Indexer, another really nice cognitive service, which lets you understand what people are saying in a video and then break that down into a transcript so you can read it, we can do real-time translation. So you really start to bring together that you might have in PDfs a big pile of information really extract knowledge and be able to have conversational apps that leverage all of this knowledge. And then finally, it's really important for you to be able to make your conversational app get better every day as you get more and more data from your customers as they use your systems more. And so we're leveraging the great analytics tools that we have in Azure focused on conversation and dialog to really give you robust, globally-deployable, super compliant telemetry and monitoring. Just quickly, we have made available recently a solution called the Virtual Assistant because so many people want their own assistant. They want their own Cortana that leverages their brand, that does what their company does. Initially we had learned about this because we have this great services team which can provide for you a solution and build a solution for you if you don't want to build that yourself. So many customers liked this, that we actually took all the work that the services team did and we also made it available open source in our GitHub repository, so that if you want to build your own solution, you can also do that really quickly. The virtual assistant incorporates speech, question and answering, languages understanding, and lets you publish your solution out through the Azure Bot Service to the channels where people talk. And then it lets you author and create your own custom skills that integrate with whatever your company does. So you can use code, you write your own custom skills, connect to your knowledge, and then of course you can also connect to third-party assistants0 if you want to do that. And then you can provide those either in the Cloud of on the edge. So these are just some of the pieces that go into building a conversational app. Again, you can get started really easily. You can have us help you build an experience. You can do something very simple, like just pointing to an FAQ that you have, and you can build a very sophisticated solution by going in and using our comprehensive tools. And just to remind you, our tools are all available, the bot builder is available open source and GitHub so you can have your developers, or developers that we can help hook you up with, use these tools to create very sophisticated tools that leverage all of the assets that you already have today, as well as build new assets to do new things to help people converse and interact with your products, and your employees and your business in a really simple way.

Video Details

Duration: 8 minutes and 54 seconds
Country:
Language: English
License: Dotsub - Standard License
Genre: None
Views: 5
Posted by: csintl on Jun 18, 2019

10893_AI_Business_School_Lili_Cheng_FINAL_V3

Caption and Translate

    Sign In/Register for Dotsub above to caption this video.