5. 12. Describe use cases for chatbots, and opening Power Virtual Agents for Teams
In this video we’re going to start looking at chatbots. So what are chat bots? What are the use cases for it? Well, you may have gone to various websites and at the bottom right hand of the screen it says hello, my name is Anna, can I help you? And you can type information in and get responses. Well that’s a chat bot and we’re going to have a look at chatbots in this course in the context of Microsoft Teams and we’re going to look at fairly elementary chatbots. So what might you use chatbots for?
Well, you could answer questions about the need for human intervention. It is available 24 hours a day, seven days a week. So you can solve perhaps common employee issues automatically. You could look up items in data verse 15 for instance. You could find people and items and you could start Power automate flaws as part of the chatbot. So let’s have a look at how we can do this in Microsoft Teams. And you do this by going to the dot dot dot here and type in Virtual Agents.
And there you can see Microsoft virtual agents quickly and easily create chat bots for teams. So I’m going to add that in. So it’s important to realize there are some differences between Power virtual agents for teams and the full Power virtual agents. This can only be used in your organization so it can’t be used with customers.
But equally the price is pretty good if you are licensed with Microsoft Office which includes teams and you’re licensed with Power virtual agents. So this has been used in the past, for instance for COVID-19, infection rate and tracking information, sales help and support issues, opening hours, store information, employee health, vacation benefits and other common employee questions for businesses. And it can be very quick to set up. So let’s set one up now. So click Start now and we need to select a team and I’m going to create my Power platform team Chatbot.
So I’ll click continue there. Now we need to give it a name. Well I’m going to give it the name of documentation and you can see the various languages that can be used. So danish. Dutch. English. French. German. And if we go down the list we can see some preview languages as well. So let’s click create. So it’s now Creating my bot and you can see after about a minute it says we’re still building it. While we work, feel free to chat with the test bot, trace your conversation and check out conversation topics. So here we can see we’ve got a test bot and so I can say hello maybe. And the computer says hi, I’m a virtual agent, I can help with whatever.
So what can I help with today? Opening hours. And I’m happy to help with opening hours. We have these opening hours. Does that answer your question? Question. And we can see that it is based on a flaw chart starting with trigger phrases for topic and then various messages until we get to an end of a conversation. There is a bit more that’s available in that. So in this video we’ve installed Power virtual agents for teams. In the next video we’re going to create our own bot.
6. 57. Create a chatbot for Microsoft Teams
Now in this video we’re going to have a look at creating gau on chatbot. So chatbots are made up of things like entities. So entities are real world subjects or concept of them. So phone number, zip or postal code, address, city, state colors, numbers, names, that sort of thing. But we don’t need to get too much into entities in the P 100 hundred exam, this is more for the PL 200. Instead we’re going to stick largely with topics and we can see we have got a few sample topics already installed and you can click on any one of these if you want to see information about it. And then you can go to the offering canvas. But instead what we’re going to do is go back to topics on the left hand side and create a new topic. So a topic is how a conversation develops.
So you can start from templates, but we can start it also from the beginning. So this is going to be my documentation 2020 or later. So we’ll give that a friendly name as well. And you can see that’s the external facing name that they will see. So how might my customers so internal customers only ask about this topic. So I’ll say where is the documentation? Documentation, where can I find a manual? And you can see try to start with five to ten diverse phrases. So the computer needs to be able to understand are when the person types in manual, for instance, he may be looking at that or she may be looking at that. It uses artificial intelligence to pass a customer’s input so he can get into the right topic. So these trigger phrases, they’re keywords questions or phrases relating to an issue. So now I’ve got this start, I’ll go to the authoring canvas.
So just saving the topic and then we will go to the authoring canvas. So they’re my trigger phrases. Next question is okay, usually when you start a conversation, you say something and someone else says something. So welcome to the documentation section for 2020 and later. Now let’s add something. We can show yet another message, we can ask a question, we can add a condition and we can call an action. So the action might be for instance, launching Power automate so you can call a floor based on your response to your chatbot.
So I’m going to ask a question, which year do you want the documentation for? And so you can see we can have multiple choice options or maybe we ask for an entity, an age, a city, a color, that sort of thing. So I’m going to give multiple choice options. So 2021-2022 you notice save the response as so it will be saved as a variable at the moment called a VAR. So I’m going to say year requested and it’s going to be text. So if it’s equal to 2020, then do something. So you can see the computer has inserted all of these four different conditions based on what I have done.
So I can have different action based on 20 20 20 21, 20, 22, 20 19. So what I’m going to do with 2020, I’m going to show a message which says the documentation can be found here and I will add a link to where the documentation is going to be found. Now obviously, I don’t have any documentation, so I’m going to use Microsoft Address. If it’s 2021, then I’m going to say the documentation can be found here and I’m going to this time use Google Address if it’s equal to 2022. So I’ll just slide along. Then my message is, I’m sorry, this documentation has not yet been released. And then if it’s 2019, I’m going to say this has been archived and we’ll call up the information.
So I’m going to save it as that. And now I’m going to create another topic. So I’ll go back to topics and this topic and you can see all we have is create from blank. This topic is going to be documentation 2019 or earlier. So archive documentation. But I’m not going to be using much in terms of a trigger phrase for this. So I’ll go to the offering canvas for this particular one and say this is the 2019 or earlier information. It could be found here that this may be out of date.
So maybe this is a chat bot that I created in an earlier year. And I just want to say here’s the information, it may or may not be current. So I’ll go to Bing. com here. Now at the end of a topic, I need to end a conversation so I can end with a survey. So the survey has just answered your question basically. So now I’ve got this. I will go back to my earlier topic of of documentation, 2020 or later. I’ll go to the offering canvas and you notice this little symbol here with the three squares can also be found here. So you can go straight to the offering canvas and there’s a few additional options.
So if it’s 2019, then what I want to do is go to another topic and I will go to the topic of documentation, 2019 or earlier. So what do we do at the end of a conversation? So I will end with a survey. Now I could put in lots of different end of conversations here as well, but I’m not going to do that. Instead what I’m going to do is I’m going to click on Add Nod so that creates this circle and then drag this circle so that it’s the same with this circle. And that just creates the one path at the end for these three.
So do that to this as well. So maybe I was continuing to ask additional information so I don’t have to replicate that lots of times. So I’m going to now save that. And in the next video, we’re going to have a look at testing at this chatbot and publishing it so that other people can use it.
