Hi there folks, and welcome to our learned live session here at Microsoft Ignite. My name is Homo and I am a cloud advocate at Microsoft focused on Power apps and Fusion developments. I'm so excited to be here, to be to be here with you. As part of this amazing conference that we're having, we'll be covering a Microsoft learn module uh titled Use Microsoft Code Copilots to Build Next Gen. copilots. You can access the module by either following the link on screen or by scanning the QR code for the by scanning the QR code and accessing the module that way. But for the next hour or so, we'll be covering the contents of that module on screen for you. Now, today I'm with my colleague Jeff Blitz. How? How are you? Hello, level. How's it over there? I'm good, I'm good. It is quite late, just after 11:30 PM, but I'm good. Great. It's interesting it's, it's 10:30 PM over here too. But it's amazing to be speaking our days and then live session and to be featuring our Microsoft Ignite. I'm Jeff and I am a cloud advocate with the growth and ecosystems team. And it's my pleasure to be back again this year to share with you on some amazing things that are happening at Microsoft Ignite. And interestingly, this year we will be talking about an amazing solution which was about at Microsoft Ignite, so stay tuned for that. But he we're not alone here. We have a couple of amazing fellows who are behind this is in the chat section and they are keeping their eyes on your questions or your comments just to make sure it goes well. I have Kim fellow who is a cloud advocate on the same team as me and we get to help people learn about technology and do the amazing stuff that they need to do. And then we also have Alicia Tapia who is a tech strategist at Microsoft, Another amazing fellow who is in the chat section to make sure that your questions are as said. Aside these amazing moderators and ourselves who are presenting, we have you all who are joining this live session. Thank you for attending Microsoft Ignite. Thank you for also joining this specific session. But he we don't want this entire session to be boring, so please be sure to use the interactive chat section. Say Hi over there to Los area joining from I'm joining from Nigeria to lose some of the amazing things that you have seen happening at Microsoft Ignite that excited about and we can't wait to dive into the session proper with you. As I had mentioned, this specifically live session is based off a module with the team creates copilots with Microsoft Copilot Studio. Don't forget to utilize the link that's being displayed on the screen AKA dot Ms. slash learn live- 2023 Triple 15-C. It will take you to that line module and you could have access to it, or you could also utilize the QR code, you know, just scan it and it'll take you directly to the land module. Would you like to talk a little bit about some of the things that are going to learn from this session? Yeah, sure. So some of the learning objectives that we have in the session, they mirror the ones from the learned module as well. First up, we're going to be spending a bit of time just speaking about Microsoft Copilot Studio. It is a brand new product which was announced yesterday. So we'll just spend some, some time just speaking about what that means and what that, what types of opportunities that opens up for us. And then we're also going to be speaking about some of the generative AI capabilities that come with that and how you can use that to create amazing Co pilots. And then we're going to have a bit of an exercise where both Jackets and I will be taking you through how to build a Co a copilot with Copilot studio especially using AI to not only generate answers from different datasets, but also to generate things like topics, phrases and responses as well. So I'm pretty excited for this module. Like we said, it is day two of Ignite. There's been an amazing events so far. We've had two amazing keynotes, um, which cover a plethora of announcements surrounding dev to dev tooling with Visual Studio, GitHub, the power platform, along with all the amazing updates with Azure and infrastructure. You know, data and of course AI. But today we're going to be introducing you to, of course Microsoft Copilots Studio. Um, this is a again a product that was announced yesterday and I love the tagline that we see on screen. Right, Your Copilots your way. And that's exactly what it is. It's an end to end conversational AI platform that lets you create and customize powerful AI powered copilots using natural language or a graphical interface for a whole range of requests from providing simple answers to common questions to resolving issues requiring complex multi turn conversations and all the other features that you've come to to to know with conversational AI. So this type of tool enables you to engage with customers and employees across industries, department roles in multiple languages, across websites, mobile apps in any channel that you may choose across the Azure Bot Framework. Now for those who might not have caught the announcements, with Copilot Studio being an end to end conversational AI platform, there are multiple ways in which you can build out these rich conversational experiences. You know, starting with building your own copilots rights, you can go ahead and create and publish a custom copilots for your business using the intuitive building experience of the Copilot studio and the large language models and and generative AI, in addition to creating your own Co copilots. What's interesting about this is that you can go ahead and customize and extend the Microsoft Copilot. So if you're interacting with copilots in any specific way, you can also use Microsoft Copilot Studio to extend that code that that copilots with your own sorts of workflows using plugins that connect to your specific data, right. And one other thing and this is something that I'm also quite um excited for is the connected platform of the Copilot Studio rights. Since Copilot Studio is essentially built on the foundation of you know of PVA and other conversational AI technologies like Azure AI Studio, Azure Cognitive Services, Bot Framework, Power platforms, AI and so much more. And lastly, no Microsoft products would be complete without its rich security and governance experiences where you have a large breadth of governance and control features that enable you to monitor usage across this platform with full visibility around all the the customized copilots that you create, the standalone from scratch copilots that you create as well as well as who is building, who is customizing across the entire environment. I don't know about you, Jeff Flip, but this is a pretty, I think, revolutionary tool to come out of Microsoft in the last few months. Yeah, it's it's really, really exciting. It's actually changing the way we did a lot of these things now, giving us the ability to actually create things on our own, things dependent on our data, things dependence on the processes, the instructions, the prompts that we give it. And that's that's really something that I, I love about the entire solution. What what something else you actually love that for? You stood out during the announcement of Microsoft. All ballots through you. It's actually what is on our screen right now in editioning to, you know, extending and customizing some of the capabilities of the original Microsoft copilots. It's this ability to build our own custom copilot as well. And this is really the focus of this learned life. And there are two main ways in which we can do that. Those two ways which we'll be covering, umm, later on in the session. The 1st is that generative AI pots of creating copilot copilots. So you can use AI to essentially make your copilots useful right off the embeds. All you have to do is give it a resource and it will use generative AI to consolidate information from those resources to answer any questions your users might have. But the other parts that I'm quite excited for is the ability to use natural language to build out the topics that you create inside of your Copilot. So topics are essentially these nodes or these workflows that your Copilot can work through because of course you can use the generative AI capabilities to create these conversations. But if you want to have more control over the type of data that is being pushed out by your whole pile pio copilots, then you are able to use natural language to build out those topics as well. And again, we will be covering this quite shortly. Yeah. And I'm also excited about that feature because for me personally, while building boards, it was a little bit difficult coming up with those topics, coming up with those notes and saying how do I connect them, How do I ensure I have the right variable for this specific don't. How do I ensure I'm connecting them properly? But hey, if you were not convenient so that you were like me, Copilot, Microsoft Copilot Studio is actually making that process super, super easy for you. Don't worry, we'll show you how it goes. But just before we dive into the demos, would you like to probably talk us through Global level what we are going to be working on during the demo section? Sure. As I briefly mentioned UH earlier on UH, a large part or a big parts or an exciting part of building Copilot with Copilot Studio is the fact that you can use something called Generative Answers or what it was previously known as conversational boosters. So essentially when you when you are creating your custom copilots, you are able to give it for example a website right? A website, either a public external facing websites, all internal websites or internal links to like things like your internal SharePoint sites. Or you you can connect to data in Azure Open AI or more recently you can upload documents. So if you have certain policies and documents that you keep internal, you can use this existing knowledge base, this knowledge base that your companies have worked so hard on and upload those knowledge bases into the bots right from the get go. And it'll use generative AI to create answers based on that data. So if I were to use the copilots in any way and if my question references something in that data, I don't have to do much to the bots. I just have to point the dots to that data set and it will try and create an and answer. And what's great about this feature, it's not just a you know one and gone right you are able to enable Multi Turn Chats over this data as well. Again a feature which we've all come to know and love when it comes to conversational AI. So we we made sure not to miss that in Copilot Studio as well. But the other parts is the generative builder, which is something that Jeffrey will be covering. Yeah, so for the generative builder section, another exciting feature. It's actually helping you to create your topics, helping you create all the notes that you need without you having to do much of the brainstorming, much of the thinking. But here's another great feature that I love most. It's actually helping you improve your topics by giving you suggestions. It also helps you create the phrases, the descriptions, the names that you would need for your topics. So all you need to do is build a prompt, stall it what you need to do, and then it's going to create everything for you. Another thing is the fact that in between these different nodes, aside the fact that you can create a note specifically for you, it can also help you edit your loads. So probably you need something to be attached, Maybe a message to be attached to a specific note, or a question to be attached to a specific node. This generative builder future would actually help modify it and make that possible for you. Within minutes you can go ahead so also your different topics and actually break your boards to life. Don't worry, during the entire period of the demo will be showing you how this works, so I think we should probably head over there and Level will show you what the Generative answer section of the Microsoft Copilot Studio looks like. Yeah, sure. And please feel free to again ask any questions that you might have in their chats. You know, we have an amazing team in the back end who are helping us onto some of those questions. So with that, we'll be moving on to our exercise where we will be building a Copilot inside of Copilot Studio, especially with the new AI capabilities. Let me just go ahead and just close this um slide deck here. Let's open up Copilot Studio. And again how we access this is just through Copilot studio.microsoft.com. Hopefully everyone can see what's on screen. Do let us know if you're having any issues there. So it's very similar if you have used PBA in the on passed, it's quite it's it's quite familiar in that sense. Again, a lot of the features and the capabilities of PBA are now part of Copilot Studio. We'll just go ahead and create a brand new Copilot there and let's give it a name. So go ahead and call it surface copilots, I'll name that ignite Um. And then as we said, um earlier on, you can, um now use a multitude of different languages, much more than we had before. So again, I'll just stick to English there. Once you set your copilot language at the beginning, you know that's it. You can't really do much after that as well when it comes to the language. And this is where the magic really happens. Right now we're creating an empty copilot, right? There's no data, there's no real conversational flow. But what we can do is we can add in a website. So what I'll do is I have the Microsoft Surface websites opened up on the side here and then with that, I'll go ahead and just paste that in. So we have a website, we have a sort of knowledge base or starting off knowledge base that we'd like to put in there and we'll go ahead and create and create our copilots again. That's all we had to do. We had to just name it, select a language and point it to an initial data source there. And this is what it looks like. Let me just close the testing pane of the of the copilots, Yeah, And we'll see. We have things like generative AI. A lot of what we've just been speaking about with the generative answers with the generative building is in this generative AI section. So we'll go into that just now. You can also create custom topics right? So apart from just using AI to create answers for you, you can go ahead and have further control with the types of topics that you're Co pilots you know works with and and this is also something brand new with the plugins, right? If you have, if you want to connect to data that is not in a website that's not sitting in a document somewhere or a SharePoint site, you may maybe you connect to the data through an API or a connector. You are able to then create a plugin that taps into that data. So when people are asking questions through your copilots, then it's able to tap into that data as well. And this is something that's been a long time coming. And of course, the big one is that you can extend any of the Microsoft copilots that your company uses, which is actually pretty cool. So what I'll do is I'll just go ahead and open up the testing panel for the Copilot. And because we have added in, that's umm, that's websites for the Surface. I will go ahead and just ask a basic question here. I'll say, what is the latest Surface Laptop, right? And go ahead and hit send. And hopefully what that should do is it'll read my question, it will look at the data sets that I gave it and it'll bring out the latest Surface Laptop, right? This is the Surface Laptop 4. I do know that there might be one later, but again, if you're looking at the at the data source, it's referencing the South African websites as we see from that globalization key there. So I'm a bit sad to say we don't have the latest Surface Laptop, but as you can tell, it is referencing the latest information. So you know that your users, your copilot users will always have the latest information from from there. What do you think about that Jeffrey, that is that is pretty. Cool. Yeah, pretty cool. I actually love the fact that you have to go through the entire process of having to update at intervals and adding new information and improve your your, your, your notes to take the correct cloud. Great messages, it does everything for you and ensures that that they are updated and that's super super interesting. One other feature that I love with with this. Yeah, one of the phrases that has been thrown with this particular feature is that it makes your chat bot instantly useful, right? All you did was you pointed it to a data store and it works straight away without you actually customizing anything, which is pretty cool. So if we click into our generative AI section here you will then see how that works. Specifically, we are able to add not just one website, but multiple websites so it can reference multiple websites to generate a single answer, which is actually something quite useful, especially when you have a large data sets or your data sets in multiple places. So in addition to public websites, you can also add your internal SharePoint sites as well. And of course in preview you also have the ability to upload documents as well. As I've said earlier, if you have policies and certain internal documents that come that have been built since out of your company, you can leverage that existing knowledge base inside of the AI tool as well. So as soon as you add more websites or you added additional documents, you just click on save and that brings up that section over there, which is actually pretty sweet. But now of course this you don't have much control over the data that comes with generative answers. That's where topics come into play. So topics enable you to add in specific items that you want the bots to speak about that might not be in these in these various knowledge bases that you have provided. And with that, we'll pass that over to Jeffrey to show us how that's done. Yeah, sure. I guess you can see my screen at this point. So UMM level had walked us through creating our Copilot. So I won't show you how to create that just so we move fast with this. Now we have created our Copilot already and now we need to create a topic for specific section of our website of our organization. We're working with a real estate organization and this refix session we need to create a topic for is people who need to actually come to learn more about the real estate. They found the specific real estate item and they want to learn more about it. They want to find more information about that. So what we're going to do is come over here and as usual if you're conversant is provincial agents, you know these are, these are topics that have already been created for you custom topics and you can head over to create yours. But this time you can do so with copilots and we're going to give us a name so it's we're going to call it book a real estate showing. Something we need to understand next is what are the details we need to collect from this specific user. Something I always like saying to people when I'm introducing Copilot, any of the copilots pictures is you always remain the pilot. Copilot is just an assistant to help improve the process. You are the pilot of the entire process, you remain the one behind the scene. Creating all of this and our transmitters was both to point different ways in which copilots would show that it's a copilot are not the pilot of this entire solution. So we need to understand the details. Like I mentioned earlier, I'm and some of the details we want to collect is the users full name. We want to also collect users e-mail address to communicate with them. We want to identify the specific property that they are interested in, and we also need to collect the date and time that they're interested in coming to that specific property to have a view of it. So we're going to head over and bring our prompts together. Remember, prompts are lines or words that you use to tell your copilot what to do, and based off those prompts, it carries out specific set of actions. So we don't come here and say just. Before you carry on, I think that's an important note to bring about when it comes to prompts, right? Just like how you had said that copilot is exactly, it's the copilot, so you are still the one in charge. When it comes to prompts, you need to be very specific in your prompts, right You need. You need to tell it exactly what you need, or else you won't get the desired outcome. Yeah. And that's very important because I've made lots of people who are like, hey, I told Copilot to do this for me, and he couldn't. And I'm like, OK, how did he explain it to Copilot? And they realized that they probably just had two words or three words of probably a few words. And Copilot was like, Oh my God, I don't understand this, but hey, let's dive in and go back to that. So we're going to collect users full name. Uh, so that's just each users your name. Next thing I'm going to do is collect users e-mail, collect the address of the property and then finally we're going to collect the date and time that they use, UH wants to actually compute that property. We're going to head over and create and let's see what's copilots would get done for us. So I'm just going to take this, get this out. Notice I'm keeping it as created trigger phrases for us. We don't have to think of what our user is going to impute and I mean Gomolemo probably can relate to this and probably those who are currently on this session that points way you are thinking to yourself, how would people you know tell my boss to carry out this specific action or activity and you have to do a lot of the thinking. Don't worry, pilot is already need that for you with tons of trigger phrases right? And now it's going to do a create a note, a question note, that's what is the users full name is going to answer the users e-mail address. But here's something we want to do in between the question node for the full name and the e-mail address once you actually think the user for providing his details his full name. So we're going to come here and if you notice, I have highlighted the question node for the e-mail address and I'm going to go over to edit with Copilot and I'm going to say, hey Copilot, please help me update um, the message to say thank you. So the name of that variable is full name. So full name variables. Umm from the previous note, Let's see. And then proceeds to axe the question. Now notice some kittens or based off the specific problems that we have written first. Notice that it is referencing the last nudge. So it's not just it didn't forget what was impacted at the beginning. It's still taking notes of what was impacted. And that's why it's a conversation, right To conversation still remembers what was imputing last and can carry out activities based off whatever information has been beaten. So come here, yeah. Just just as we like do like a quick, uh, backtrack. I think it's important to note that the fact that it's not just there at the beginning to create the topic, you can still work with Copilot to edit the topic as well and refine it and make it better. Like if you don't, if you don't know how to add certain nodes, or if you if you don't know how to use variables like how you're doing now. You can use copilots to assist you in that in that feature as well, which is pretty cool. So you essentially using copilots to build a copilot, which is a bit, it's a bit weird, but it's it's, it's, I'm cool nonetheless. Yeah, so I'm just going to put that text again, I snow. So again, I think this is the, IT goes back to what we were saying earlier that we need to be quite specific with what we are saying, umm, so that it understands, but it again it might might not always get it right and then you can just go ahead and fix that in the in the, umm canvas. Yeah, so let's come back here. Let's see what specific note it's actually selected. So if you notice it's actually coming down here, so just going to take this down and bring it back up, that's that's again and I just had to copy it so I don't have to get it again. It's actually one of the things with copilots Who right you won't always get the exact answer every time. Again, it is AI, so you need to work with this ambit and then if it doesn't work out, you can just go ahead and edit it as well. So what we're going to do is also includes the comma here and we can see the image message variation. And just so because of let's try it one more time and let's see OK, so let's just come over here and imputes the message variation. And I wanted to do was to actually as you can see from the last note that was created here to say thank you very much full name string. So I'm just going to copy this out and coming here to the real section and I'll that in myself. And just it it just goes to I'm sure right the the copilot does not know everything, right you are. You still need to have that sort of control as to what it's what it outputs, how it works. So you can't just fully rely on its to do everything that you wanted to do exactly. It's just it's just a tool to help you get to that destination faster. Yeah. So what I'm going to do is save this and then something else we also spoke about is the Adaptive Card feature. Basically the Adaptive Card feature is really interesting stuff that adds all of these things at where all of the details that has been in SO and it includes all of them for you and you can get over the, the users can head over and see all of the different details that they imputed. So what I'm going to do next is come over here updates and I would say summarize the information that's being tired by the user. Ohh, so let me just collect and Adaptive Card and what it's going to do is that it's going to create and Adaptive Cards, so flex and error. So let's go over and refresh this and then we'll just head over and do it again. Great. So let's come over here. I'll take this out and what I'm going to do is impute this again. So let me use this specific prompt that has been created here. Still, your requests didn't go through, so let's go over the entire process again and see what's exactly the challenge might have been. So let's build this again. Let's users one here. E-mail address of property, dates and time of the showing and let's just create this again. Great, so let's have this section done again. Umm. So again, I think as you going through this this part again I think it's important to note that when you want to make a change to a specific node, just make sure that you select it first. As you can see, like as you were going through it, you selected the node you want to make a change to and then you try to edit it. Otherwise, the copilot does not have the context of whatever node you're trying to edit. Yeah, that's another valid point. So let's go over and click on this white section. So ensure no node is selected and then it it's with Copilot. And let's try out our Adaptive Card feature again and hopefully this time it works out. And I mean, and then luckily we have the learned module, so if it doesn't we'll just show it. I I guess we yeah, the demo what with us tonight. Yeah, So what I could do is actually just in case this doesn't work. I had tried it and it worked out. Let's see. OK, and the adaptive card section has worked this time puree. So notice that it's empty. So what I'm going to do is come here and I can see the property is section. So let's go back properties and notice this is the specific formula it gives me. I have a rose and this was something else I wanted to show you. Copilot can make mistakes and it will make mistakes. This would not go through. If I actually try saving this and testing out, it's not going to work because of those errors. So what did from the last vote was that I edited the formula. Now I'm going to be staying the new formula. So just come in here and paste it. This is the new board tech created. This is how the formula should actually look like. You know, taking out all of those columns that showed that the title, the variables were strings. So let me just take it back and you can see that we had these columns in front and behind that actually showed that this was a string instead of it being there's something different. Answering questions in the chats and someone said that you know what, sometimes it's just easier to go ahead and do it yourself instead of using the Copilot. But it's really about, you know, making sure that the the the gates to entry is easy for everyone to amuse, right. So if you are one who has more familiar with the features of Copilot Studio, which what is now Copilot Studio, then go ahead and you can and it's as you yourself. But the the purpose of using natural language is to make sure that everyone has the ability to build out these copilots based on the data that they work with every day. And I think that is really the beauty behind what this product does. So then we just go back again and this is the correct code. If you go over to the land module you'll find that this has been updated. So you can just come over here, copy it and paste it the correct formula. So I'm going to come here and I should see that my formula is good to go. So just save it and then give it some time for it to get saved. So now at this point, let's actually test it out to see how it works. So I'm over, I've saved this properly. I'll come over here, that's the copilot and I can head over here and say book a real estate showing. So just appeared to me full name. You are already aware of my full name, my e-mail address, and just foods. For the sake of the demo, they addressed 49 OK because this should work. So the data time for 70 since 3:00, 1200 PM and you can see that the outdoor active card has HOMO. It's now requesting me to review the information provided in the Adaptive Card. And this just shows that you are indeed the pilots and the situation. Yes, exactly does to really really that simple. That's easy for us to get started or get done with this parts. Remember from what you have said, hope pilot is just that assistance is there to help you out. You remain the pilot, you are the one in charge of the solution. So like my my my partner had already mentioned, you have to be conversant with what you are building. You have to be conversant with the technology you are using and then you support copilots along the way while it gets to build the solution and and actually it, like you mentioned, understanding it's having an idea. If I didn't have an idea what the specific error was about, I wouldn't have been able to do it. So have an idea definitely of what the prompt type. I'll just have a basic understanding of the different parts of it and it's easy for you to make whatever changes that need to be made. Yeah, you can go ahead and extend this further. Again connecting to those custom plugins that tap into your data or a connector that you created as well, you know, save it in some sort of data store from this point. But using those AI, those generative AI features to 1 to create generative answers from websites. And then in turn to have more control over some of the topics that you create through the generative builder or the copilots to edit a Copilot UM, that's just some of the things that you can accomplish with Microsoft Copilot Studio. So with that, we'll just go back to our slides and what we will do next. We've done a bit of talking and a bunch of demo. Um, we'll just do a quick Q&A. Our knowledge check just to reference some of the things that we have done just to make sure that we are on par with what we've shown so far. So the first question, I do know that you will have links that will be pasted into the chats or some sort of poll that you can use as well. And we are seeing those polls live as well. So it'll be nice to see what types of answers you guys put in. But Jeff, Blitz, um, what is the primary purpose of copilots in Microsoft Copilot Studio? Um, we have 4 possible answers. Yeah, you know, to automate websites searches. That's an interesting one. It could be possible. It sounds good BI to enhance the chat bots appearance. I'm also not a bad shot. So that's that's actually quite good. Um, see to help with bot creation by using natural language. You know, Sounds familiar I don't know. Or to deploy bots to multiple platforms. We're getting a bunch of votes in. So far about 93% of voters are going for C What do you think? So First off, web service website searches, that's actually the goal of copilots from what you mentioned, yes, it goes over to the website and finds information, but that's not the primary purpose. So a is not the answer for the user interface, the appearance, that's not also the answer. So, so we're looking at between C&D, let's give some time to see how many persons will foresee and how many persons go for D and go over and check out what the answers be. Yeah, We have multiple people voting for all the answers. But for now, first place is see with about 94% and we have second place going to D with about 4%. So I think we're in a safe place. We can go ahead and reveal that the answer is indeed see, right. It's to help with the bot creation using natural language and that's exactly what Jeff Let just showed us right now using that natural language Copilot text to help you build a Co pilot and Copilot Studio, right. The majority of you got that right. So that is great. Let's go ahead to question two as well. Yeah, question 2. Which feature in Microsoft Copilot Studio allows the bot to present information from multiple sources without the need for creating topics, right? This is also one that can be a bit tricky. We have Generative Responses, Copilot Builder, Conversation Booster, Sorry, and Azure AI Bots Service. A lot of these things serve the copilots in different ways. A lot of these things might not even exist. I don't know if that is a big a big clue there, but let's see what the responses are coming with through what are you thinking? They're definitely. So it's definitely, uh, we haven't spoken about Azure AI Bot Service, so I don't think that's an answer. The is no part of it. Umm, I'm looking at A, B or C Let's see what people were actually going for. This is actually an interesting one because I'm looking at the votes coming in and they seems to be the majority of people are voting for B while we have second place going to now see and then we have third place going to a conversation booster. Now I don't know if it's maybe I'll folks without explaining it properly I don't know or maybe there was a bit of confusion with some of the wording. You know brand new products, you know brand new features but in this case it's the minority, the minority of the people 15 percents of viewers voted for A which is conversation booster. Now I think we can understand why there is a bit of a confusion here because it does use generative AI to create those, um, generative Answers, right? I know that's very similar to Generative Responses, but it's Generative Answers and Generative Answers uses that to boost your conversations through the URLs and knowledge bases that you provide for that call pilots. That's the answer for question 2. Let's go. Do you have one more? Let's go for one more. Yes, we do. The last question and this is how do you test your bots in Microsoft Copilot Studio? Or rather, how do we test your Copilot in Microsoft Copilot Studio? Yeah. Yeah. Yeah, yeah. It's gonna take a while to get used to. Yeah. Again, full answers a deploy the bot to Azure, right? You spoke very briefly about the certain channels that you have available to publish your your copilot, so maybe that's an option. We can select the Test your parts button in the lower left corner of the screen. We were testing the bots quite regularly throughout the building, especially in your part of the demo, so that is a good shout as well. See by turning on the track between topics toggle. We did not cover this in this section, but this is something that can happen inside of Copilot Studio. But is it the answer to this question We don't know. Or by publishing the bots to a test site? How? How? How do we test our Copilots javelin? Hmm, I I think actually telling people the answer would not make this one, so I would like to keep my mouth sealed. So until we get lots of responses. Yeah, if you're looking at the votes, the majority, about 91% of us are voting for B, while the second highest votes are going to D, which is quite interesting. Umm, should we just pull them out of their misery? Yeah, I think we should, I mean it's understandable for a lot of people to actually go for deep ball. By the time we're going for D we're actually a little bit moving out of the environment of Microsoft Copilot Studio was actually saying within the studio you're not going out, you just testing it and that was be the lower left corner of the screen. Remember that point where I had to impute the users questions? What is I wanted to say? And then it went ahead to actually give me the responses, right? I think it's been pretty good working through all of this. We had about 87% of viewers voting for B, so that's a good chunk. And here, whether you got it right or wrong, something I'd love to mention is you can actually go back to the module and get acquainted with all of these things. Don't be scared of failing. Don't be scared of not getting this right. You just go back and continue learning Gomolemo. Would you like to do a quick summary of some of the things we've spoken about in this session? Yeah, yeah. So again, just going back to what we spoke about at the beginning, we briefly spoke about what is Copilot Studio and how it can help help you to, one, create your own custom Copilot Copilots or extend and customize the currents Microsoft copilots that some of you have access to. And then using some of those generative AI features, namely Generative Builder, to create topics inside of the copilots as well. Or Generative Answers to use existing knowledge bases either in the form of documents or websites or SharePoint sites to generate answers for your users using those copilots. And we also went through a demo to build out these copilots using these AI features as well. Again, we didn't make a satisfactory sacrifice to our demo gods. How? However, we do hope you saw the flow of how that works out, and we hope that you also visit the learn module to just try out some of these features for yourself. This is also part of a larger learning module that speaks about copilots in other parts of other Microsoft products as well, which is pretty cool and in the spirit of Copilot Studio. Some of you were asking a bunch of these questions in chat, stuff like do we have a community page or or is there current docs that we can look through as well, Please, Here are some of the resources around Copilot Studio. You know, fresh off the the unpressed, we have our announcement blog. We have a demo. If you don't have access to Copilot Studio yet, you can see how those generative AI features work in our demo link. We have the Sizzle video just for you to see some of those highlights as well. Again, the docs explaining all the various parts of the Copilot Studio, how you can go ahead and customize your own copilots as well. And I hope you guys try Microsoft Copilot Studio once you leave this session. Uh, Jeff, let what is next? Yeah, there are other sessions that are part of the Learn Life series, be sure to check them out. This is actually focused on Microsoft Defender be deploying in my defender for endpoint environments and this is going to be another interesting one where you get to see how this comes to life. So be sure to also try out this Learn Live session and enjoy the remaining parts of Microsoft goodnight. Thank you. Cheers.
