/ Insights / View Recording: No Code, No Problem: Steer the AI-Powered App Journey with Copilot Studio Insights View Recording: No Code, No Problem: Steer the AI-Powered App Journey with Copilot Studio June 12, 2025Think AI-driven app development is only for coders? Think again.In this hands-on session, you’ll see how Microsoft Copilot Studio empowers professionals across roles—whether in business, IT, or operations—to turn natural language into fully functional, data-connected applications. No code required. No waiting on dev queues.You’ll learn how to:Craft effective prompts that translate real business needs into appsQuickly model and structure data in Microsoft DataverseRapidly prototype and test solutions with real usersWhether you’re defining processes, solving problems, or accelerating delivery, this session puts you in control of the AI-powered tools reshaping how work gets done. Transcription Collapsed Transcription Expanded Brian Haydin 0:12 Fantastic. Well hello everybody. Thanks for joining us. I’m Brian Hayden. I’m a solution architect at concurrency. We’re gonna talk a little bit about low code solutions. I just got back a couple weeks ago from Microsoft build and one of the things that I was just absolutely blown away with was some of the new tooling that’s come out in the the power platform. Copilot Studio’s something that I’ve talked about a few times. In fact, actually this this talk that we’re gonna do today is. Serve our Rif off of the new stuff that’s out there. And so thanks a lot for joining us. And let’s go ahead and dive in. Like I said, my name’s Brian Hayden. I’m a solution architect. I’m a outdoor enthusiast, so a lot of my talks and discussions use analogies that you know I bring from my, you know, outdoor world, into the technology world. Fun little fact about me is. Is that I’m a twin. Who’s next? Who wants to introduce themselves? I got a couple people with me. Ajay Ravi 1:21 I can go down to next. So my name is Ajay Devi and I’m a senior engineer working with concurrency for a little bit. Over the last six years to I work mainly on the dynamics 365 power platform and low code solution helping organization streamline and build. Digital solutions. So yeah, nice looking forward to this webinar and kind of like go over some of the cool copilot studio and look good stuff. Brian Haydin 1:50 And Eli? Eli Rodriguez 1:51 Yeah. Hey everyone. My name is Eli Delriguez. I’ve been a system engineer concurrency for 5 1/2 years. I think also working a lot on the power platform stuff. So seeing all these new things coming out with copilot studio and how the power platform is evolving with it, it’s very exciting. So we’re really, really excited to show everything that is going on with y’all. Little fun fact, I moved from Wisconsin to Texas. That’s why you can see that in McAllen, TX. Image I couldn’t deal with the weather, so I’m back to to the heat. Sorry to meet you. Brian Haydin 2:22 And heat you have down there, that’s for sure. Eli Rodriguez 2:25 Yes, yes, I do. Ajay Ravi 2:25 Mm hmm. Brian Haydin 2:26 All right. So I like being a little bit interactive. I don’t have a pole to throw up here, but maybe drop a a comment on what your role is. This this talk is really geared towards business analysts and so if anybody remembers the movie office space, you know, I I guess I’m gonna ask you like, you know, what is it that you say that you do here? Here you know for a job, if you go to the next slide, you know the question I have for everybody is you know, what is your take on AI technologies and how it’s going to affect your job? You know some people you know think that, you know, AI is going to take my job and for business analysts, I think that’s especially, you know, at the top of their mind. Like you know, if I can just use an agent or. Take copilot notes like what? What value do I have? So next slide. The answer’s no. So I think that, you know, building software you know is kind of like, you know, trekking through the woods trying to find the right path. You know, I’ve got all these different tools that, you know, compasses, GPS. But at the end of the day, I need a good. Guide and so I think business analysts are uniquely positioned to be able to be a trail guide you know. For the developers to to actually do the work, and in the case of today, maybe even just the developer out of it, so next slide. What makes you as a business analyst? You know good at your job and I would say that, you know, go back to the office space. You know, the guys screaming. You know, I’m a people person. You know your soft skills are the key. Your ability to communicate with business. You know people that are doing work and. And find out like how is it that you’re doing it? Why are you doing it this way? Critical thinking of the the problem at hand and coming up with a a really well reasoned solution for them. These are the things that make business analysts. You know fantastic. I would even say, you know, human centered design comes into play with a lot of the UX people that I’ve worked with. You know they think about, you know, the problem space and how we actually go about solving it. And most of the people that I’ve worked with. Within this space, you know are always learning new things. I can think of a couple people that have worked with a concurrency that were always pushing, you know the boundaries. If you took a look at our ado instance, our Azure Dev OP instance, there are so many plug. We had to start taking them out because they were crushing us, so they’re always looking for new tools, ways to get things done and so that kind of added adaptability. You know something that’s really treasured in in your space. Next slide. So what are some of the top trends in 2025? You know, around AI. I mean, generative AI has been around for a couple of years now. But we’re starting to see those technologies being applied to product ownership, backlog management, things of that nature. I see a lot of product owners using generative AI to craft well formulated. Did well reasoned, well documented user stories. You know, it can often think of some of the edge cases or acceptance criteria. You know that people you know don’t. You know, don’t always think about. I can give you really insightful reasoning around that data and analytics. You know, behavior skills. This is becoming increasingly important, important. Having the ability to collaborate not just with your coworkers, but. Coming prepared to a meeting with, you know, a lot of thoughts. You know, that’s one way that I use copilot is help me prepare for this meeting. What are the top three questions I need to ask? You know, those are all really good. Really good use cases for this, so next slide. And this is kind of a poll opportunity. You know, feel free to drop something in the Q&A, but when I do this talk. Talk in person. I kinda ask people you know for their thoughts and perspectives. How are you using AI today as part of the the requirements analysis? As part of the you know process of building software in the sdlc. So I mean some of the low hanging fruit things that I do every day are you know when I have a meeting with a client, you know I get the automatic summary. But I don’t just. Stop there. I always ask it for follow up questions and intuition and insights around the meeting, and I use those to gather my requirements. So a lot of my requirements analysis comes directly from individuals. Couple presentations I’ve done recently. I can say here are the facts that I heard and actually call out who the individual was, what the meeting, you know time and date was that we talked about it. And it actually is kind of cool that I can like, hey, Jim, you’re the one that said this. On 525, when we met to talk about this, I already mentioned that, you know, backlog grooming generative AI is really fantastic. For that testers are using are using generative AI to create test data. We’ve got a demo a little bit later around the treasure hunt and just to kind of flash forward to that I I used ChatGPT to like go and find a bunch of these, you know, hints and clues and. Uh put together a list of 100 things. Would it take me awhile to put together 100 different test cases that actually are real, but I could use, you know, AI to do some of that? But any. Any questions or any comments in the chat? I’m not seeing anybody drop anything in, but I’d love to see some of your ideas how you’re using it as well. I don’t want to the next slide. But all this brings up an important point that there’s still actually, there’s not just still, but there’s more of a trust gap in AI. Than than there was a year ago. And what’s really interesting to me is listening to a podcast a couple of days ago that this is somewhat a uniquely American thing and that the trust gap is in in the US is actually growing, whereas it’s shrinking in like the. Eu and. You know APAC, but taking a look at this graph, you can see that over time, less people are willing to use, you know, rely on AI systems, maybe because they’re seeing hallucinations, maybe because they have found some of the, you know, like, anomalies that happen when you’re using. It a lot of people are seeing AI systems, as you know that they’re worried about them. That’s gone from 49% to 62%. In two years. And that trend is continuing to go, you know, in kind of a negative direction. But you know AI, if you use it properly, you know we’re gonna see that it isn’t necessarily replacing us. It’s helping us get, you know, to our destination a little bit faster. So next slide. So I’m gonna take a step back from like, you know, just sort of framing this up. And I wanna frame up an application so. So I’m a software developer engineer by by trade. Been doing that for over 20 years and you know when people ask me like to build an application, my mind immediately goes to something with a screen and an interface and you know and I start thinking about low code. Like copilot studio. As you know, are those really applications? I really had a hard time kinda framing that in my head and so I wanna take you on the journey. And. And help you get to the place where I did, which is, you know, rethinking what an application really is. But at the end of the day, an application has, you know, several different components that are important to call out. Typically it’s got a user interface, so some way for the user interact with it. It does some sort of logic. It might be that we’re processing medical claims. It might be that we are collecting information. It might be if you have a a website with shopping portal, you know you can only buy two of these things. Or if I buy this then I can buy that and then I have to be able to store that data. So I’ve got data. You know data storage and persistence. And then at the end of the day, most applications aren’t just there to collect information. They actually do something with it. So there’s some output. Sometimes I work with companies in like manufacturing, so that output could be a physical output. I’m moving something on a conveyor. Or it could be that the output is. I get something shipped to my house. You know that I purchased on a website, so next slide challenge that I you know wanna propose to everybody is you know what really is an application. And in the early days. You know, we had these. You know green screen applications and over, you know in the 50s, sixties and 70s and as the 80s kind of rolled around in the 90s, we started having client server applications. That, you know, got really goofy with, you know, too many buttons on the screen. And then most recently, we’ve been doing things in SAS platforms and web access, you know. And so again, Michael thought when you say go build me an application. Is something that looks like what you see on the screen today, so I’ll ask question give you a second to pause about this next screen. Are these applications? The events there, there we go. So if you take a look at these things, are they applications? And I would argue that they all are applications. They typically collect information a user has to interact with them and they’re going to help them achieve some goal. Whether that might be tracking, you know, my fit. On my watch or asking ECHO to play a new song. These are different modalities that you know I’m working. With a computer system or an application in order to achieve my goals next screen. So you know, thinking back about that, how can we look at non traditional ways that people do things as an application, think of somebody who’s doing like a sculptor, a sculpture, that art that artisan, that person that’s doing the woodworking, you know, on the screen. Maybe they, you know, they don’t want to work with their hands so much, but they want to use the creativity. So we turn that artisan and manually working on the woodwork into. To Acnc router and we can create these designs using. You know some AI agent and allow people to that haven’t put thousands or 2000 or 3000 hours worth of into their craft and actually turning into something beautiful. So at the end of the day, what is an application? And I came up with a single simple line. You know description of it. It’s just a self-contained program. Designed to perform certain tasks, whatever that might be. To next screen. And using that as kind of a basis for the definition, what are what are applications gonna look like in 10 years? That’s a long way out with with technology going as fast as it is today, but I think that they’re still gonna have the same major components. We’re gonna feed information into it through some sort of human interface. Whether that might be my voice, whether that might be you. Know signals coming in on the smartwatch. But it’s gonna have some sort of interaction with, you know with. Think with humans and it’s gonna be able to do things on our behalf and not only just do things on our behalf, but do it intelligently. Think of what we’re actually trying to accomplish, and maybe you know, help us get through our destination faster. So next screen. Now this is where kind of copilot studio comes into play, and I think I’m gonna turn this over to one of the other gentlemen to start framing up the conversation in in copilot studio. But you know, from a definition standpoint, copilot studio was rolled out recently as really more or less the chatbot. I guess you would say eliyah J would you disagree? Yeah. So, but even with a chatbot, you know, with an AI driven chatbot, it’s a human interaction. Somebody’s typing into it. They’re doing it without buttons and without the controls. But they’re using their natural language, and now we’re voice enabling it. And so we have inputs. We have outputs. We have intelligence and we have actions. So I’m gonna turn it over to these guys to talk about what copilot studio is. And give you a short little demo. Take it away guys. Ajay Ravi 16:16 Thanks, Brian for that, the interest. So we can do like just quickly go over like what copilot studio is and when few things and then you know we’ll kind of like go over some of the demo items and test you can kind of like build as well. So Microsoft introduced copilot studio a couple of years ago, doing an Ignite session. So it is a low code, no code platform that. Allows you to build intelligent chatbot agents and other like the term is getting updated every now and then. So it’s called agents these days. So you can build these intelligent agents using very minimal code or like even not no code I would say. But it doesn’t mean that you cannot use pro code or there are definitely features and functionality that allow you to include and inject that pro code functionality as well. But. Pretty much it’s a look good environment and it will allow you to build. Intelligent agents to perform specific task and activities for you and also you can utilize the generative AI functionality without applying any. Occurred or building any models and things like that so you can use the LLM model behind all the Gopal studio to kind of like you said Gen. AI functionalities as well. So again. It was there before, you know, Microsoft obviously introduced copilot studio. There was like a power virtual agent within power platform, but it did not have the generative AI functionality but with the whole introduction of copilot studio rebranding and then including that Gen. AI functionality. There are so many additional capabilities as well. So few key things to kind of like consider and you know kind of like litigation about copilot studio. So Microsoft recently introduced a multi model conversational experience, which is really. Cool and neat. So the whole concept is that you can actually build multiple agents to perform specific tasks rather than building 1 agent for doing like, you know 10 or 15 different things. You can build individual copilots for all certain things and then kind of like integrate everything together so it. Is currently in preview and I think it’s a really great feature that would definitely be helpful as well for all the different business functionalities. And another cool thing. About Copilot studio is that without applying any code or anything so you can connect with several knowledge source and data sources. So I can quickly show you a couple of examples shortly, but you know, being a business analyst or from a business side of things, you can actually start building some of these agents with few clicks and we’ll quickly show how easy it is to kind of like start. That, but again, you do have the ability to make it. By adding like you know all those additional data sources and everything as well. And again, our sensor is closely connected with power platform. There are over 1000 plus of plugins and operable connectors are available, so that lets you connect to your different business applications like Salesforce or work Day, Dynamics 365’s or service now and all those platforms as well. OK so. Just wanted to quickly share like how to get started with, you know, Arch copilot studio. I’m going to, you know, show a great example of how to do this so. Right now I am in the copilot studio portal, so once you come to the copilot studio portal, you can click on this new agent functionality and that will allow you or like you know you can say copilot inside COPE studio. So with this shared. And our national language, you can actually start building Echo Palette studio chatbot within like you know I would say like few minutes. So what? Is I am going to, you know, ask for a copilot studio that allows you to, you know, give best practice for project management. So I’m going to mention something like. Project management. So so I just provide like a high level details on what kind of agent I’m looking for. So on the right hand side, you can see that with every instruction I add, it is kind of like adding some instructions and everything on the back end as well, so. What copilot did was like it recommended, like you know. Are you OK with project management guru as your name still is going to say? Search. So now you can see that the name got updated to management kuru and then it is asking like you know what kind of like functionalities or things should the scope alert should focus on. So I’m going to like, you know, quickly mention that provide. Perspectives. I’m keeping it simple for now, and then I’m also going to provide a sample as well. That is like a external data source as well, so I will go ahead and do that and I’m just going to provide like a data source. So it is pretty much confirming like is there any other modifications? So I’m going to save this. And then what I’m going to do is once everything has been added so you can see that all those prompts details, everything has been added in here and by clicking on create it, you can actually create a copilot studio with the basic inscription, and then you can defin. Fine tune all those details as well. So this will probably take a second. So now you can see that the description, the general inscription and all those details have been added in here. And then if I go to the knowledge section so you can see that the website which I provided was just like the public website that has been added autom. As well. So this is how quickly you can actually start spinning up studio. And then definitely you can add additional components add in functionality all along the way as well. I understand you came out of mood and I think you want to like add there. OK, sounds good. So. Brian Haydin 23:18 Like the we just call out really quick that this was all just like you interacting with it as like from a business sense there was no real technical, you know questions, everything was built for you in the back end. And using just you know a good business sense or business requirement? And I think it’s important just to call that out that this didn’t require a developer skill set in order to get this up and running. Right. Ajay Ravi 23:51 Thank you for that, Eric. So yeah, that is pretty much how quickly again actually start copilot studio and you know include all those additional functionality. Then you can add different functionalities, agents and everything associated to that as well. So. What we’re going to do is there will be kind of like going over this quick scenario. It’s a function type of thing which you guys can do as well based on, you know, you know your availability. So I’m going to quickly go over like a scenario where we can actually build a fun. Activity as well. So the goal for this particular scenario is that based on the location which you provide the copilot or using the AI functionalities. Allen ADVER functionality, along with some of the data sources. It will pretty much give you some hints. And then you know, you can kind of like get more which particular location or a place the copilot or the AI is talking about? So I think it would be like a fun excel which you guys can to. So what we have done for this particular scenario is we have used dataverse as the data. Knowledge source. So we have showed like about 200 ish different clues. Along with like a particular location. So we have used database have the knowledge source and then use chatbot for kind of like all the interactive purpose. So you can kind of like ask different. You know, I would like iclu for Chicago or like Boston, you know that kind of thing. And then it will accordingly look for all those databases and then provide you some clues. So that way you can kind of like guess and play that so. Pretty much generative AI is kind of like comes into play and gives you all those hints. And then you can kind of like, you know, play around with this and in future if you want to add additional context like external sites like Wikipedia and all those things, you can definitely include that as well. Brian Haydin 25:55 And the cool part about this is that we’re this isn’t just a a chat interface to some information. This is intended to demonstrate actually doing work you know, with copilot Studio, right? Ajay Ravi 26:16 And then just to kind of like quickly go over some of the copilot studio prompts and details. Useful so we have added like few functionality details along with the different components that you need to configure if you want to set up the scope studio so all these things will be added to the copilot studio as prompts and restrictions and then the user flow would be. A person with open copilot studio and you and then kind of like ours, you know, provide the current location. And then accordingly. A script or like you know a hint will be given to the user and then they can. Kind of like get that as well. So here’s a quick example of like how the conversation would look like. So I’ll just for the sake of time, I’m going to quickly go like, you know, some what I was kind of like doing earlier today. So I just mentioned I wanted to play an adventure game, and the copilot pretty much asked for, like, you know, the location. So I mentioned Chicago. So it’s not like giving looking at the database and then provided me like some. Clues for the first one so you know, I kind of like guessed that. Because I’m in Chicago, so I was able to like guess our cloud kit and then, you know, you can kind of like lay around and then kind of interact with the copilot studio to play with this cool game. And just to kind of like, give you a quick. Overview So what we did was this is pretty much like an example of how we stored the data. So you can see that there’s a clue text which we have added in here. Along with the location and what the answer would be and the hint. So we have connected this knowledge source to all the copilot studio so you can see that if I go to the knowledge part I can see the game clue. So this is pretty much where we have stored different. Hints and information about the game and then we are using all the power automate or like an agent flow which is the latest name. So we are using that for kind of like playing and interacting with this agent. So you can see that it’ll be like pretty much like a basic chat type experience. Fair you can pretty much mention that I want to like play game and then provide the location and it starts giving you different examples of the hints and then you can accordingly guess as well. So it really need an easy way to kind of like maybe like a fun activity type of thing, which you can kind of like do all based on your time and then you know. Do that. So going back to the SharePoint, so another thing which we want to kind of like mention regarding copilot studio is that so copilot studio allows you to build all those shared experience. But in addition to that there are there is this agent flow which is previously like power. Automate so you can utilize agent flow to interact with your business system and get all that information as well. So again, you can build all these agents without much, you know, spending much time by the simple chat type experience. So you can see that. Here’s an example where you also have the copilot within an agent flow, so you can go ahead and you know, kind of like mention something like. I want to create a flow to get the list of all database records from the team CLU table. So by just kind. Of mentioning few text information. There you can actually you know, the copilot would take a first stab at it and then build the rough flow. And then depending on what kind of changes you need to make, you can accordingly do that as well. So these are really, you know nice way to get started by. You know, even if you don’t have much technical experience or like development skill set, you can use these copilots to get started and then. Expand and then make necessary changes as needed. So again, a great way to you know, interact with your business system and get that information back to the copilot studio. Bring over a little bit around the knowledge source. So again, these are some of the key or the common knowledge sources that are available in copilot studio. So if you want to connect to a public website or an internal SharePoint or dataverse dynamics 365 or service now you can actually you know, go to that knowledge part in copilot studio and integrate that. So that way whenever a user interacts or asks questions. No to the copilot studio agent, then it will pretty much look at these knowledge sources that we have provided and then provide the next region AI results. So this way it’s not just looking at the LM, but again it is looking at the different documents and information that you are providing. So that way the answers and you know response is purely based on the data which you have provided. So again, it’s a pretty and you can also upload documents like guides and everything directly off from. From SharePoint or SA documents. So again an easy way to give knowledge and knowledge to copilot studio to allow it to kind of like look at those knowledge and then get an appropriate results. Brian Haydin 31:54 We got a list of the ones already in the Advanced tab. I noticed SAP was in the advanced tab that came out just recently. So these are just the easy ones, but there’s definitely a lot of other knowledge sources that you can tie into as well. Ajay Ravi 32:08 Mm hmm. Thank you for bringing that one up. So again, like I mentioned, so these are like the basic ones, but if you go to the advanced ones, so you can see. All these one and then there are always the power automate or the agent flow that you can kind of like integrate and then kind of. Interact with those systems as well. Thank you for bringing that one up, Brian. OK. So just want to briefly touch on the whole multi agent orchestration as well. So this was something which Microsoft introduced a couple of months back. Maybe like during build. So the whole purpose is that you can actually build multiple agents for specific task rather than one agent doing everything from end to end. So that will allow you the different businesses or users to build their own agents and kind of all interact accordingly. So again, this is a whole new concept and I think this would be really beneficial so that we each agent has his dedicated or designated or task or activity that it needs to perform. So you know it will be easy to kind of like manage. So that way every you can build all those different agent for specific business related tasks. Again, Yep, it’ll be currently in preview mode, so in the next month or so you can actually start playing around. And see some of the. School arbiters etc. And just kind of like briefly go over the channels part. So once you build these koalas studio agent, you can actually deploy these channels to different location. So there’s a demo website where you can kind of like use for testing purposes and if you want to deploy these copilot studio into teams or Microsoft 365 Copilot, you can actually do that like pretty quickly for Microsoft 365 Copilot, you might have to. Work with your team’s administrator to deploy that, but. Again, these are some of the internal channels and SharePoint is a new one. Which Microsoft introduced recently, and in addition to that, there are all these additional channels where you can deploy these copilot in a web app, Native app, Facebook and all those different places. And if you want to integrate with a customer engagement hub where you need to hand off. The copilot studio chat to a live agent. You do have the ability to do that by integrating with. Customer service, you know, a sales force service now and all those custom engagement hub as well. So with this you can kind of like build Escalade to do and then deploy to different channel based on your difference. And just to kind of like add, SharePoint and WhatsApp are the two new additions which Microsoft recently made. So this will allow you to, you know, deploy the scope alerts to SharePoint, SharePoint and WhatsApp as well. OK. So I am going to hand over to Eli for some additional functionalities around planner and then a quick demo so. Eli Rodriguez 35:21 Yeah. Thank you, Ajay. So as we move into a world where where people, business users, developers are working more and more with AI agents. The way that we think about how we build these apps is is really starting to shift right, and Microsoft Power apps is playing a big role in how this mentality shifts, how we work with the agents, how we use them day-to-day, how we managed all Sol. At an enterprise scale and because of that, that’s what we have these three releases that we have on our screen. So we have plans in power apps which are now under GA general availability. It’s given not just developers, but makers. Makers could be anyone in your company could be a CCE 0. Could be a President, a manager, business analyst and developers. A specific team that is going to help you build your solutions. You can go from just an idea all the way to an end solution. In a matter of minutes, we also have the new agent feed in power app. So it’s still in public preview, but it does give users a very intuitive in your experience to manage it managed how they collaborate with agents. So it’s gonna give you, like sort of like a hub where you are gonna be able to see all the agent actions that are being taken in your apps. You we’re gonna get also real time guidance on how to complete specific items and then you’re also gonna be able to automate some work with some intelligent actions within. This agent feed hub. We also have how the managed property is extending its world class governance, security and operational control, which is already being used by mill series around the world. So if we have developers that use development tools like cursor or Visual Studio code, they’re going to be able to deploy their code first apps into the power platform, and they’re going to be able to gain all the benefits that come with the security, management, governance and all. The deployment features. Ers. Can we move to the next screen please? So in the previous slide, we talked a little bit about plant designer, right? It was just like a little quick intro, but I want to go a little bit more into detail because I think it’s something that is going to make planning and building solutions a lot more easier for makers in your environment or organization. So what is the plan designer? Right. The plan designer is basically an AI powered sidekick that you have in your solution that is going to help you build a solution just from writing something in natural language. So you don’t have to start from scratch. You can start with an idea. Type your needs what you want the process to do what you want the app to do. You can even throw in images if you have a very very you know. A drawing of your data model or of a float that you want to have. You can add them to your prompt and then what the designer will do is it’ll start building the solution for you. So it’ll go from the user stories, the user personas. The requirements. The tables you’re going to need suggest applications that you may need all the way to power. Automate flows, power pages, and even copilot agents. So it’s really giving you a really nice head start. For your whole project your whole solution, but it’s tailored to specifically just what you need. So I’m going to go ahead and take a little bit of control of the screen right now. So I can show you in real time how it all comes together. Just one second. Are we able to see my screen? I cannot see team so thank you. Brian Haydin 39:06 Yep, I. Eli Rodriguez 39:08 Thank you for the confirmation. Brian Haydin 39:08 Yep, I see you. Eli Rodriguez 39:10 Appreciate it. So we have a a business case right now, and I’m gonna paste it right here. You can access. Your your planner from your maker portal either homepage or you can access it through the plans section. And by clicking new plan. So here what I’ll I’m gonna describe in a business problem and I’m gonna. I have a business problem right here that I’m gonna copy and paste. So this is what we have right now. I need a solution that allows employees to submit submit reimbursement requests. The app should let employees submit new and B existing requests and let managers who approve, deny or send them back for review. What I’m going to do here before clicking generate. I’m also going to include preview features, so you can also see how it generates processes for us, so I’m going ahead and click generate. You see how he’s starting to think through everything that I gave in and start building some user requirements. Some user personas on the right, you can see that is generating some processes. In here you’ll be able to see all the agents that are working currently with me so. I have the requirements agent which is built in this section. The process agent build in my processes the data agent which is going to determine which tables and things I need and then the solutions agent is going to be. These are the power apps that I think you’re gonna need for this solution. So this makes building your solution very interactive. You can also have several authors in this same session. The only difference is that only one of them can be edited. This plan at a time, so I believe you can have up to 100 people Co authoring this. Plan, but again, it’s first in first out. So the first one that gets in is the one that is going to have access to be able to edit. This, but everybody else can be able to see the changes in real time. So now that I have. My requirements and my processes I can actually look into what it’s telling me that it it believes that I need so right now I see a submit and manage reimbursement request process. So I’m going to go ahead and click view process. I’m going to hide this for a little bit and then zoom out a little bit. And you can see how it says you can. This could be your process. I’m going to start desmission. Fill the form. Submit the form. View the request and so on, right? It gives me a specific flow that it thinks I will need. For this solution I can edit things from here if I want to. So if I want to add another note, I can add it. Here I can select the role that is assigned to it and what the task will be. And then on. My on my document. You’ll see that it has the business problem laid out for me, but it’s a purpose of the plan and then it’s telling me also, what are the user requirements here? I can give feedback on whether or not I like this information or not. I can select if this looks good and I can move forward or if there’s things that I want to change in my requirements. So I’m gonna go ahead and click edit and you’ll be able to see that. I can also ask here if I wanna add more personas if I wanna change items where I can do that manually as well. So if I click on my menu here, I can add a new user role or I can just go ahead and delete this role. I can also add requirements or delete requirements, so in this case, for me this looks good, so I’m gonna go ahead and click. That’s good. So now what it’s doing? Is that the data model agent is generating tables that I want to have here. So you’ll see that it’s generating a table employee reimbursement request status. The manager and expense type. I can also go ahead and click here show details and it’ll give me my model that I can see how data is being connected based on what the agent is telling me. I can make changes manually either through. Working on this section or I can ask. My agent as well to do things for me. So in this case, I don’t think I need request status as a separate table, so I’m just going to go ahead and remove it. And then I’m gonna ask to add it as a column in my reimbursement request, so add status. As a choice column. To reimburse. Ment request table. So it’s currently working on it and you see that now I have my status options hit here. Another neat thing is if I click here on view data, it already also has mockup data for me. So it’s really easy for me to be able to know what I’m doing as I plan my solution to get a real grasp of what my solution will look like. Once I start building. So in this case, we’re gonna. Say that we’re good to go on this table, so we’re just gonna go ahead and go back. And I’m gonna say that this looks good. So now the solution agent is thinking through what I have here thinking through the requirements, thinking through the processes, thinking through the data model and saying this is what I think you need as far as technology. So it’s telling me that I need a canvas app for the for the employees to be able to submit and edit their reimbursement. That I need a model driven app. For managers, you need a flow for the notifications and then a an agent. To be able to do this through a copilot agent. So in this case, I’m gonna go ahead and click edit and you can. I’m gonna do a natural language prompt and say remove manager or remove model driven app. Everything. Needs to be in a single campus app. Go ahead and click next. And now you see that I have. It has made updates for me and now I only have my employee reimbursement app, but I still maintain my notification and my agent. I’m gonna go ahead and say I want to keep these changes. This looks good. And now we’re ready to go. So I I can leave this place and just save my plan or I can even start and actually create the tables if I am done with my data model. So for example, let’s say that I’m. I’m good to go. I’m going ahead and create my tables, so I’m going ahead and save tables and now it’s telling me to save it in a solution. So the planner is solution aware, which is great. So I’m going to go ahead and say yes, save this as my reimbursement request system solution. Let’s click in advance. You can select an existing solution or you can also change your publisher. You click save. And he’s starting to save a solution. It’ll create the tables for me as well. Right now the planner only works with dataverse, so your environment if you want to use the planner, needs to have a dataverse database installed. So I’m. I’m not sure if other sources are gonna come. I would hope so that other sources are gonna come, but for now, dataverse is what is currently supported. You can see the data agent is currently saving my tables. Sometimes it’s a little bit faster, sometimes it’s a little bit slower, but at the end of the day it will create my tables and you’re going to be able to see they’re going to go from Gray to green once they’re created. You see? Now they’re green, meaning that my tables are created. And it can take it a step further if I want to. So I can also even create my app. It’ll give me a basic app based on my requirements. So if I were to click here, I can choose to either create it, I can choose to replace it. Linux is an app that I already have in my environment where I can convert it to a model driven app. So in this case, let’s go ahead and also hover over here. You’ll see the user rows are being used. For this application it’s using my employee, my manager and is using my tables so I’m create. It’ll take me. To the studio. It’s loading on my tables, loading the screens. And just like that, I have a canvas app that is using the data that I just created and it’s giving me some basic screens to start working on. So all I have to do right now is just save and publish and make it available for my users to work on. If I were to go back. Into my solutions. You see that I have my solution here. And I can see everything that I’ve created so far. So I have my tables. I haven’t chosen to create the flows, but if I were had created the flows they would be there. And you can also see that I have my plan as well. So I hope that this demo was able to give you all a nice feeling of the things that you can do with the planner and how easy it is to make your ideas a a reality. Once you’re working on the planner for power platform. I’m going to send it back to Ajay. Thanks Ajay. Ajay Ravi 49:59 Thank you for that, Eli. OK. Hope everyone can see my screen, so just going to quickly go over like some of the additional things regarding. How you can potentially use like pro code along with the low code. So even though today’s session was mainly focusing around other local solution and copilot, studio and power apps and planners, so it doesn’t mean that you cannot use the pro code side of things. So Microsoft have introduced. Do you score interpreter which you can actually use for bringing some of the Python powered logic and everything into the code studio as well? So these are some of the things that Microsoft is heavily investing and then adding new new functionality. I mean those one of my little mentioned that they’re like so many changes that something that happened last week have been updated this week because Microsoft is constantly pushing new UPD. Every now and then. So again, even though it is a low code, it doesn’t mean that you cannot use some of the development experience and some of the pro code functionality as well. And they’re going to add Brian, yeah. Brian Haydin 51:14 Yeah. Yeah. We didn’t get into. I’m not sure if you’re you bringing up the MCP stuff and everything as well. I’m not sure what other slides you have here, but this is really fantastic. I’m. I’m gonna be doing a talk. I think in a couple of weeks about the convergence between low code and pro code and diving a little bit deeper into some of these concepts. But you know, I’ve always thought of using rectwall for the job as being important and more and more I’m leveraging these low code tools. But I get stuck, you know, and there’s just certain things that it’s not capable of doing or not not capable of doing. Well, this is a really good example of like being able to give a nice graph without having to do a round trip to something like fabric or power BI and give the, you know, give users intuitive insights in the information based on like requests that they had so. This is a really cool feature that I look forward to playing with. Ajay Ravi 52:14 Thank you for that, Brian. And also wanted to quickly go over like some of the model support functionality that are going to be coming soon or in power platform and copilot studio so. With these task specific models that will be available in Arch Copilot studio, you can actually use some of these models for specific like for example if you are from legal and if you want to like review some of the legal agreements, you can actually use one of these. Star specific models to review and compare and then make sure that you get the appropriate feedback and then also even generate some of those things. So again, some of these are in preview, but really you know great features and really looking forward to it. OK, so I think that was pretty much it. And Brian you into like briefly touch on the speech? Brian Haydin 53:12 So a lot of good questions in the meeting chat here too as well in Q&A. So first and foremost, I just want to say if you want to talk a little bit more about this with me or Eli or Ajay, we’d love to connect with you, you know, for a one-on-one session. So fill out the survey link that Paige dropped in in the chat, and we’ll get back to you and reach out to you. So if you want, you know, if you want to talk about copilot studio. Agents, multi agents, MCP, all that kind of good stuff. I’m really passionate about this technology. Another thing that we do pretty well here at concurrency is just the M365. I think. Chris, you had a a question about integrating this into teams. Everything that we’re that we showed you today is about building applications in the power platform and a natural channel. To deploy those is inside of Microsoft Teams. We can create agents that you can use inside of M365, copilot and and then we can help you. With user adoption, making sure that people are getting the value out of it as well. But before we even, you know, get started, a lot of organizations, we wanna make sure that your organization has the alignment from the executive team. And so we we offer you know works. For your executive team to answer their questions, what does AI mean? What are other companies doing in in this space? Making sure that they’re on board before you start to bring these solutions to the organization. Getting that executive. Team aligned with that is is a critical first step and then often will get ideas out of the executive team. You know, just by the questions that they ask. What do you think? Is this capable? Can we do this? Can we do that? So really good way to build momentum within the organization. Fill out that survey link will reach out to you and get that follow up session schedule. But other than that, we’ll stick around monitoring the chat here for a minute or two. But thanks everybody for showing up. I really enjoyed it and I love giving you stocks. Eli Rodriguez 55:19 Thank you everyone.