Build Conversational AI Experience using Watson Assistant

In this lecture, we will describe the example use case, we will be using for designing, and implementing the Conversational AI solution using Watson Assistant

In this lecture, we will provide step by step instructions on how to set-up the sand-box.   This includes four major activities, namely:

1. Set up IBM Cloud Account

2. Provision Watson Assistant

3. Setup your project workspace

4. Walk-through of sample project using Watson Assistant

In this lecture, we will cover Conversational AI concepts like Utterances and Intents

In this lecture, we will cover "Entities" - key concepts or objects of interest found in a given conversation

In this lecture, we will introduce the concept of Context - in a given conversation session or across session.

In this lecture, we will introduce the concept "Dialog" in a conversation.

In this lecture, we will introduce concept of Short tail or long tail questions

So far, we covered the concepts behind Conversational AI. Now we will pivot towards developing the Conversational AI using Watson Assistant.

Here we will introduce to you the overall Assistant and its skills. In this section, we will outline the organization of Assistant and Skills in Watson Assistant and will get you started with action and search skills. In the next two sections, we will cover dialog skills. So, what are these skills and how they work for an assistant? Let’s use this video to define some of these terms and explore the architecture of an assistant.

As explained earlier, An actions skill contains actions that represent the tasks you want your assistant to help your customers with.  In this lecture, we help you use an actions skill to build your first conversation associated with COVID vaccination query

First came the website and then came the FAQ’s, or frequently asked questions. Most websites are fairly extensive in their knowledge. However something led to the evolution of FAQ's. What was it? Despite placing all the knowledge about product and services, marketers were perplexed that they could not break the tide of the users. And they eventually found that it was a lot more easier for a user to come and ask an employee a question rather than to walk through second, third or fourth layer of a website.

Conversational AI offers a mechanism to not only place those FAQ's in a much more readily accessible may, but it is also paving the way to create effect FAQs using natural language processing.

In this lecture, we will show you how you can build a generalized. FAQ capability using Watson discovery and then enable it through Watson Assistant for a chatbot to respond to its users.

In this lecture, we will cover sample examples of key conversational AI  concepts for our COVID use case and how to create / train and test these Conversational AI key concepts into Watson Assistant. 

We will cover design for following conversational AI concepts in Watson Assistant

• Utterances

• Intents

• Entities

• Context variables

In this lecture, we will teach you how to build, train and test key concepts in Watson Assistant

In this lecture, we will cover Task 1 out of 6 Tasks, to design Level 2 - Basic Dialog with 3 Intents – We will provide step by step instructions to demonstrate this task 1 in this lecture

In this lecture, we will cover Task 2 out of 5 Tasks to Build/Test Level 2 – Basic Dialog with 3 Intents created in last lecture

In this lecture, we will cover Task 3 out of 6 Tasks to  Design Level 3: Dialog with 5 Intents

In this lecture, we will cover Task 4 out of 6 Tasks to Build/Test Level 2 – Basic Dialog with 5 Intents created in last lecture

In this lecture, we will demonstrate Tasks 5 and 6 - How to design, build, train and test all dialogs for our chosen use case

This lecture describes various ways of integrating IBM Watson Assistant with other systems and show cases two design patterns.  It lays the foundation for the remaining lectures in this section.

Conversational AI is an evolving and complex topic. While we can easily develop a maturity level 2 model, most real applications are at level 3 and level 4. This section will provide you glimpses of complex architecture decisions and related alternatives. The objective of this section is to introduce a number of these topics. They are by no means representing an exhaustive treatment of the topics.

In this lecture, we are summarizing the course contents

In this lecture, we are providing future courses in various AI related topics