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


0 (0 Likes / 0 Dislikes)
[AI Business School] [Introduction to AI technology for business leaders [Introduction to the Microsoft AI approach] >>David: In any enterprise, you will do three things with AI. The first one, you will create intelligent applications [David Carmona | Microsoft, General Manager, AI Marketing] by bringing AI to every application. The second one, you will optimize, or even redefine your business processes. And the third one, you will empower every employee to work better with AI. Let's see the main technologies that are commonly used in each of them. First, let's start with bringing AI to every application. A very common approach to fuse AI into applications is to use pre-built AI services. Instead of creating your AI model from scratch, you can reuse an existing model that has been trained for a particular scenario already. This model is then encapsulated in a cloud service that is very easy to call from your application. Typical scenarios that are available as pre-built AI services are, for example, computer vision, like image classification, object detection, OCR, et cetera. Also, areas like speech, with things like text to speech, or speech to text, or speaker recognition. Or things like language, with sentiment analysis, translation, or language understanding. Or knowledge, with things like anti-destruction, search, or question and answer generation. In the case of Microsoft, these services are called Azure Cognitive Services. And the product lead, Lance Olson, will talk about that in way more detail in a video later in this model. The second approach is to create entirely new applications that are only possible with AI, based on conversational experiences. These conversational agents can change the way that your employees and your customers interact with your company. Just like every company creating a corporate website, like 20 years ago, companies are now creating their own digital agents. And just like your corporate website, it would be representing your brand. And it will be exposing your business services. To create one of these conversational experiences, you will use a conversational AI platform. With a conversational AI platform, you can create the core, the essence of your conversational agent, and expose it through multiple channels, like, for example, other applications, like a website, or a mobile application, or third party messaging channels, like Skype, or Facebook Messenger, or Teams, Slack, or any other. You can even integrate with other digital assistants, like Alexa or Google Assistant, or Cortana. These two things, pre-built AI services and conversational AI platforms, are often very integrated. When you build your digital agent, you will usually need pre-built AI services to make it smart. For example, you will use things like language understanding, or speech to text, to power your digital agent. Lili Cheng will cover conversational AI in more detail later in this model, as well. The second big pillar was to bring AI to every business process. And you can divide that in two. You have horizontal processes, which means that they are common across many industries. For example, marketing processes, or sales processes, or HR. And on the other side, you have vertical processes, which are very unique to your industry and to your organization. These are the processes that make your company differentiated. The ones that make your company, like, your company. For the first one, the horizontal processes, you may want to use a packaged SaaS solution, which are ready to be deployed. And it will only require things like the integration with your existing systems or customization. And that will give you a great time to market. Later in this model, Steve Guggenheimer will show you the ones that are provided by Microsoft for processes like sales, marketing, or customer service. On the other side of the spectrum, for very unique processes in your organization, you may want to create a custom AI solution from scratch. This will help you differentiate your company even more in the areas where you already have expertise, IT, or data assets. For those solutions, you will need an AI platform that covers things like preparing the data, training the machine learning model, deploying that model, and monitoring that model in production. And you want all of that to be as easy and as productive as possible so your team can be successful creating those solutions from scratch. Matt Winkler will cover later in this model that scenario in way more detail. Finally, you also have things in the middle. These are the accelerators that help you bootstrap your custom solutions. For example, you can use pre-trained models as a starting point, or templates for specific scenarios. In our case, we have the Azure AI Gallery, for example, that contains hundreds of templates and models that you can customize for your specific scenario. The last pillar was bringing AI to every employee. There are three key concepts associated with that. The first step to democratize AI for every employee is to democratize knowledge. Enterprises have huge amounts of data. But it's not accessible to employees to leverage. And because AI needs data, if you want to bring AI to every employee, the first thing that you need to do is to bring the data to them. That data should be easy to use and easy to understand. And it should include both structured data, such as operational business data, but also unstructured data, like documents. Pablo Castro will talk later in this model how we can use AI to also extract insights from these unstructured documents, as well. Once employees have access to that knowledge, they can use AI on top of it to do something meaningful. An easy way of doing that is by leveraging AI in the applications they already use. For example, Excel brings now AI capabilities that are built in, so you can get insights on the data. Another example could be Power BI, which allows business users to apply pre-built AI model very easily on top of the data. And finally, the ultimate realization of bringing AI to every employee would be to enable them, not only to consume AI, but also to create new AI models. This may sound difficult, but it's not. For example, the same Power BI, that I mentioned before, enables users to create net new models from scratch with a technology called "automated machine learning." Without automated machine learning, the user just needs to provide the training data, and the system will find the right machine learning model, and it will train it for you. All of that, automatically. Imagine the power of that. You can enable every employee to create their own models for their work, turning them, effectively, into citizen data scientists. With this, we finish this video. For the rest of this model, you will hear about the products in Microsoft that are aligned with each of the areas that we cover, directly from its leaders. I hope you enjoyed. Thank you very much.

Video Details

Duration: 7 minutes and 58 seconds
Language: English
License: Dotsub - Standard License
Genre: None
Views: 7
Posted by: csintl on Jun 18, 2019


Caption and Translate

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