Insights View Recording: Accelerating Dev: Harnessing the Power of GitHub Copilot X, Copilot Studio, and Azure AI Studio

View Recording: Accelerating Dev: Harnessing the Power of GitHub Copilot X, Copilot Studio, and Azure AI Studio

Discover the essential toolkit that propels your developers to new heights, featuring the dynamic synergy of GitHub Copilot X, Azure AI Studio, and Microsoft Copilot Studio. Uncover the strategies and insights your team needs to stay ahead in the fast-paced world of software development. Join us for a knowledge-packed session that equips you with the tools and know-how to supercharge your development process and drive innovation within your organization.

Transcription Collapsed

0:0:0.0 –> 0:0:1.780 Nathan Lasnoski We’ll get started in about 2 minutes here. 0:2:4.940 –> 0:2:5.430 Nathan Lasnoski OK. 0:2:5.480 –> 0:2:6.480 Nathan Lasnoski I think we’re going to get started. 0:2:8.320 –> 0:2:9.430 Nathan Lasnoski Welcome everyone. 0:2:9.760 –> 0:2:13.180 Nathan Lasnoski We are going to talk today about how to build faster with AI. 0:2:14.10 –> 0:2:16.890 Nathan Lasnoski We’re going to cover three different topic areas. 0:2:16.900 –> 0:2:23.400 Nathan Lasnoski We’re going to cover the opportunity to be able to take advantage of core development with GitHub copilot. 0:2:23.730 –> 0:2:38.250 Nathan Lasnoski We’re going to talk about building AI enabled medium code applications with copilot studio, and we’ll talk a little bit about Azure AI studio and how it relates to the general architecture and then how to take action on it, introduce ourselves a little bit. 0:2:38.260 –> 0:2:39.300 Nathan Lasnoski My name is Nathan Sonoski. 0:2:39.310 –> 0:2:47.570 Nathan Lasnoski I’m concurrency CTO been involved with concurrency for about 22 years now and spend a lot of time helping customers take advantage of AI. 0:2:47.610 –> 0:2:49.300 Nathan Lasnoski So really excited to have this conversation. 0:2:49.620 –> 0:2:50.930 Nathan Lasnoski And then we also have Brian Hayden. 0:2:50.940 –> 0:2:52.340 Nathan Lasnoski Brian, you want just introduce yourself a bit. 0:2:52.990 –> 0:2:53.380 Brian Haydin Yeah. 0:2:53.390 –> 0:2:54.20 Brian Haydin Hi Nathan. 0:2:54.30 –> 0:3:5.630 Brian Haydin I haven’t been here 22 years, but I’ve been here six or seven years and I’m a solution architect that concurrency working a lot in the app development AI space with our customers. 0:3:6.620 –> 0:3:7.50 Nathan Lasnoski Awesome. 0:3:7.480 –> 0:3:8.830 Nathan Lasnoski And then we also have a Jay. 0:3:8.840 –> 0:3:10.620 Nathan Lasnoski Ajay, do you wanna introduce yourself for a minute? 0:3:11.360 –> 0:3:11.790 Ajay Ravi Yep. 0:3:11.880 –> 0:3:12.390 Ajay Ravi Hi, everyone. 0:3:12.400 –> 0:3:12.930 Ajay Ravi Good morning. 0:3:13.20 –> 0:3:23.990 Ajay Ravi So my name is Ajay and I work mainly on the local nocode side of like the development platform such as Dynamics 365, power platform and I’m working with concurrency for almost the past five years. 0:3:24.190 –> 0:3:26.100 Ajay Ravi So yeah, really looking forward to this webinar. 0:3:26.870 –> 0:3:27.520 Nathan Lasnoski Awesome. 0:3:27.980 –> 0:3:36.100 Nathan Lasnoski And she’s gonna cover some of the copilot studio content and you get a demo, some things that I hadn’t seen before until he showed it to me this morning. 0:3:36.200 –> 0:3:39.400 Nathan Lasnoski So I’m really excited for you to be able to see the same things. 0:3:40.170 –> 0:3:41.500 Nathan Lasnoski So these are our three topics. 0:3:41.510 –> 0:3:50.480 Nathan Lasnoski We’re gonna spend the majority of the time on the first two and try to take it slow, make sure we go through some content that I think is gonna be impactful to you. 0:3:51.130 –> 0:3:55.880 Nathan Lasnoski One thing I want to call out as you get started is make sure you put your questions in the chat. 0:3:55.890 –> 0:3:57.60 Nathan Lasnoski We have three of us on this call. 0:3:57.460 –> 0:4:3.110 Nathan Lasnoski We want to make sure we’re answering your questions as we go through the content and if we have to save it to the end, we will. 0:4:3.120 –> 0:4:6.470 Nathan Lasnoski We can do that too, but as you have questions, throw them in the chat. 0:4:6.480 –> 0:4:7.220 Nathan Lasnoski Let’s get it going. 0:4:7.230 –> 0:4:12.950 Nathan Lasnoski Let’s get the conversation flowing and make sure that we’re helping you as you have those that have popped up. 0:4:14.150 –> 0:4:26.560 Nathan Lasnoski OK, so first topic, we’re going to talk about is GitHub copilots and this is a technology that is one that was one of the more earlier released AI tools, but has come a long way. 0:4:26.650 –> 0:4:32.80 Nathan Lasnoski And we’re gonna talk a lot about some of the new capabilities that have been released and are coming over the next year. 0:4:32.930 –> 0:4:46.630 Nathan Lasnoski So as you think about the maturation of development efficiency, I want you to take a good look at this slide and what you can see is a band that indicates current state capabilities and then future state capabilities. 0:4:46.840 –> 0:5:0.380 Nathan Lasnoski And as you can with any AI solution, there’s this transition that’s happening from a developer, a developer with tools, a developer starting to delegate to an AI agent, and then eventually getting to a completely autonomous agent. 0:5:0.560 –> 0:5:8.360 Nathan Lasnoski So on the far left hand side you have a developer that is building applications with no additional tooling other than their IDE. 0:5:8.520 –> 0:5:20.420 Nathan Lasnoski You can even think about it like when you first learn to develop, you’re sitting in your high school class using Pascal or gwbasic or something crazy like that, and you’re just building and you’re you’re taking advantage of what’s there. But they’re. 0:5:20.430 –> 0:5:21.850 Nathan Lasnoski You don’t have these extra tools like. 0:5:21.860 –> 0:5:24.530 Nathan Lasnoski Intellisense wasn’t even there, and you’re like, what do I do with myself? 0:5:24.680 –> 0:5:42.730 Nathan Lasnoski God, I don’t even have Intellisense, but you can see how as you move into having something like Intellisense, you immediately have a productivity boost just because you don’t have to worry about spelling quite as much, you’re able to tab out to the things that are important to you as you’re writing, and that’s where a lot of developers have sort of left off today. 0:5:42.740 –> 0:5:47.630 Nathan Lasnoski They have a set of tools that they take advantage of like Intellisense, but they haven’t necessarily made that next jump. 0:5:48.240 –> 0:5:54.30 Nathan Lasnoski So as you’re making that next jump, you’re seeing a developer start to leverage things like chat, GPT personas. 0:5:54.440 –> 0:6:2.830 Nathan Lasnoski One of our developers, for example, has prepared chat GPT personas with a whole write up that says you are a X. 0:6:2.840 –> 0:6:31.340 Nathan Lasnoski You are a cloud architect that answers questions about what the structure of this application as relates to an Azure cloud architecture should look like, and you’ll use that as a way to like delegate tasks over to that chat GPT persona that gets combined with some of the things we’re going to talk about today with copilot in the developer experience indicating that you want it to tab out a function or a data structure or a starting point for you to build a certain type of application. 0:6:32.150 –> 0:6:37.750 Nathan Lasnoski But then what got released last year was even the ability to chat with your your. 0:6:39.800 –> 0:6:43.810 Nathan Lasnoski Your code base to chat with your your question that you’re trying to build. 0:6:43.820 –> 0:6:44.190 Nathan Lasnoski So what? 0:6:44.200 –> 0:6:45.530 Nathan Lasnoski What should this look like? 0:6:45.620 –> 0:6:48.150 Nathan Lasnoski Can you can you build this structure for me? 0:6:48.210 –> 0:6:55.890 Nathan Lasnoski Give me the code I need to be able to develop this component of the application and you’re even starting to get to a point where you can kind of delegate in that sense. 0:6:56.180 –> 0:7:2.240 Nathan Lasnoski And then what doesn’t exist yet is this spot of truly autonomous agent development. 0:7:3.140 –> 0:7:5.70 Nathan Lasnoski We’re kind of at the precipice of that. 0:7:5.160 –> 0:7:13.50 Nathan Lasnoski We’re at the precipice of where we’re like, we’re starting to delegate aspects of our attention over to a chat agent to do some work for us. 0:7:13.110 –> 0:7:14.240 Nathan Lasnoski Look at it and use it. 0:7:14.730 –> 0:7:27.580 Nathan Lasnoski We haven’t yet quite gotten to the point where the agent developer, the AI agent, truly functions as like the intern that I have like the person I’m giving tasks to that’s able to actually build those tasks that’s coming. 0:7:27.590 –> 0:7:31.180 Nathan Lasnoski That’s not here yet, but we need to build the skills that get us to that point. 0:7:31.410 –> 0:7:35.800 Nathan Lasnoski So Brian, what sticks out to you when you think about some of these different stages? 0:7:36.760 –> 0:7:37.0 Brian Haydin Yeah. 0:7:37.40 –> 0:7:40.290 Brian Haydin So, uh, what really sticks out with me? 0:7:40.300 –> 0:7:52.230 Brian Haydin Is this this story about how as a developer, when I was spending 6 or 8 hours a day like writing code in that second slot, you know developer with Intellisense? 0:7:52.240 –> 0:7:53.600 Brian Haydin I mean I would use Jetbrains. 0:7:54.160 –> 0:8:0.800 Brian Haydin Uh, you know very religiously in terms of accelerating my development, you know, I wanna create a new method. 0:8:0.810 –> 0:8:2.710 Brian Haydin I wanna do you know some refactoring. 0:8:2.720 –> 0:8:4.120 Brian Haydin It was just a great tool. 0:8:4.410 –> 0:8:13.650 Brian Haydin Save me a ton of time once I had it in my toolbox like I wouldn’t let go, and if I went to an organization that didn’t have it, I I buy my own license. 0:8:14.430 –> 0:8:22.820 Brian Haydin Uh, and now, today, having used copilot, especially with the chat interface, I feel like this is the new tool. 0:8:23.130 –> 0:8:35.820 Brian Haydin This is what I mean in order to get my job done quickly, even with the chat GPT, there’s just the latency of like going between your web browser and your IDE and stuff all that stuff becomes like seamless. 0:8:35.830 –> 0:8:38.440 Brian Haydin It’s integrated within the copilot chat. 0:8:38.610 –> 0:8:54.800 Brian Haydin I’m really looking forward to the future state and we’re going to talk a little bit about that today in terms of the copilot enterprise features that are that are just about to be released, but I can’t imagine doing any development work without using something like copilot going forward. 0:8:59.880 –> 0:9:0.90 Nathan Lasnoski So. 0:9:2.70 –> 0:9:14.700 Nathan Lasnoski One of the things that I’ve been thinking about is how copilot, in a similar sense to what she just said, Brian is essentially a force multiplier to every developer, essentially like I’m I’m a developer Pre Intellisense. 0:9:14.750 –> 0:9:17.60 Nathan Lasnoski I’m a developer post Intellisense. 0:9:17.540 –> 0:9:24.680 Nathan Lasnoski I’m really excited to point where we start thinking about it like a side developer that has an ability to do X tasks. 0:9:24.690 –> 0:9:30.990 Nathan Lasnoski So one analogy I was using is like my older son can clean the garage. 0:9:31.130 –> 0:9:36.710 Nathan Lasnoski I can say Ethan, go clean the garage and he’ll go do it like he’ll take everything out of the garage. 0:9:36.810 –> 0:9:37.870 Nathan Lasnoski He’ll sweep it out. 0:9:38.100 –> 0:9:41.130 Nathan Lasnoski He’ll put everything back and it organized fashion and it’ll look great. 0:9:41.360 –> 0:9:43.950 Nathan Lasnoski My younger son, he can’t really do that. 0:9:43.960 –> 0:9:46.390 Nathan Lasnoski He kind of falls into like a jello on the floor. 0:9:46.400 –> 0:9:52.360 Nathan Lasnoski When I asked him to clean the job garage, he’s like he can’t compute all the tasks necessary to be able to do that. 0:9:52.370 –> 0:9:57.50 Nathan Lasnoski But if I said sweep the garage, that’s a task that he can do right now. 0:9:57.60 –> 0:9:59.890 Nathan Lasnoski Copilot has tasks that can do well. 0:10:0.140 –> 0:10:6.580 Nathan Lasnoski Has tasks that can’t quite do yet, and we’re still kind of discovering even like, what are all those tasks that can do well? 0:10:6.590 –> 0:10:8.100 Nathan Lasnoski What are all the tasks that can’t do? 0:10:8.220 –> 0:10:14.760 Nathan Lasnoski But it’s gonna keep aging and maturing to a point where it can be like the anti outsourcing tool. 0:10:14.770 –> 0:10:18.500 Nathan Lasnoski So like right now it’s kind of like a side developer and a really basic sense. 0:10:19.580 –> 0:10:25.430 Nathan Lasnoski I can do certain things, tell it what you want to write and have it create it, and then that’s your first draft. 0:10:25.440 –> 0:10:28.630 Nathan Lasnoski Like one of our colleagues will say, never write a draft again right. 0:10:28.820 –> 0:10:37.590 Nathan Lasnoski Like use this as a starting point to be able to do certain types of work, and it performs best when you provide clear outlines around what you have to do. 0:10:37.600 –> 0:10:41.840 Nathan Lasnoski Like my older son, he can clean the garage, but like copilot is not quite there yet. 0:10:41.850 –> 0:10:45.30 Nathan Lasnoski So I’ll say Benedict go sweep the garage. 0:10:45.40 –> 0:10:45.490 Nathan Lasnoski Do this. 0:10:45.500 –> 0:10:49.950 Nathan Lasnoski Come back to me when you’re done and I’ll give you a new job that’s sort of where copilot is at right now. 0:10:49.960 –> 0:10:51.620 Nathan Lasnoski Like I’m gonna give you a task to do. 0:10:51.730 –> 0:10:53.640 Nathan Lasnoski I’m gonna put my energy somewhere. 0:10:53.680 –> 0:10:55.80 Nathan Lasnoski I’m gonna go back to that task. 0:10:55.90 –> 0:10:57.40 Nathan Lasnoski Pull that in and have it completed. 0:10:58.800 –> 0:11:3.250 Nathan Lasnoski Where I think it can get to and this is like totally the future is. 0:11:3.300 –> 0:11:20.620 Nathan Lasnoski I think it can get to a point where it is like a developer that has a whole set of other assets available to it like pseudo human individuals like intelligent agents that are autonomous that you give to do things in a similar sense. 0:11:20.630 –> 0:11:25.140 Nathan Lasnoski The way we think about outsourcing today, like I give it a I give it a set of requirements. 0:11:25.150 –> 0:11:30.0 Nathan Lasnoski It goes out and does those I come back in the morning and I take that and I incorporate into my application. 0:11:30.230 –> 0:11:31.240 Nathan Lasnoski It’s some point. 0:11:31.250 –> 0:11:33.140 Nathan Lasnoski It’s gonna get kindness to that spot. 0:11:33.150 –> 0:11:41.380 Nathan Lasnoski It’s not there yet, but that’s where we have to start sort of preparing ourselves as AI enabled developers in order to be able to function in that particular space. 0:11:41.390 –> 0:11:47.480 Nathan Lasnoski Like Brian, what’s been your experience of it as a force multiplier in like a very like, practical sense? 0:11:48.150 –> 0:11:50.0 Brian Haydin Well, I mean, you’re getting to the next slide. 0:11:50.80 –> 0:11:52.230 Brian Haydin We’re going to talk about like, you know what it actually does. 0:11:52.720 –> 0:11:53.140 Nathan Lasnoski There you go. 0:11:52.780 –> 0:12:0.610 Brian Haydin Ohh but, but you know my experience is that it it it’s probably about 50% it it increases my productivity about 50%. 0:12:1.20 –> 0:12:7.480 Brian Haydin I’m not wasting a lot of time, you know, doing remedial kind of programming tasks. 0:12:8.70 –> 0:12:21.280 Brian Haydin I don’t have to skeleton out like unit tests when I’m working on it and and overall you know I can just focus on the more creative aspects of what I’m trying to accomplish. 0:12:21.290 –> 0:12:35.590 Brian Haydin Like, what is the actual problem that I’m trying to solve here rather than, you know, all these other like nuanced things that I’m dealing with and the previous slide, I think like, you know, one of the things I wanted to talk about and we, yeah, alright, fine. 0:12:35.730 –> 0:12:41.160 Brian Haydin Bringing back here, but you kind of skipped a little bit over the 9010 rule. 0:12:41.650 –> 0:12:41.940 Nathan Lasnoski Umm. 0:12:41.590 –> 0:12:45.680 Brian Haydin Yeah, I would say like you know, it’s not perfect. 0:12:46.110 –> 0:12:53.450 Brian Haydin I wouldn’t, you know, try to throw some really complex business logic at it and say, write me a method that does, you know, these 18 things. 0:12:54.50 –> 0:12:55.860 Brian Haydin Uh, but it’s really good. 0:12:55.950 –> 0:12:57.880 Brian Haydin And you will have to review it. 0:12:58.290 –> 0:13:13.740 Brian Haydin Another thing to note is that there are differences between the chat GPT models that are being used in conjunction with personas and the copilot, so you know some developers like to run it by both and see which gets a better result. 0:13:14.210 –> 0:13:29.890 Brian Haydin That’s just going to improve and especially when you can start to ground your your repos to two, copilot can actually gets get results that are relevant to you know, your internal development repos. 0:13:30.120 –> 0:13:31.850 Brian Haydin It’s just gonna get even better. 0:13:32.540 –> 0:13:36.170 Brian Haydin So that’s and then we can go back and talk a little bit about the results. 0:13:36.580 –> 0:13:40.370 Brian Haydin So the overall. 0:13:41.210 –> 0:13:48.700 Brian Haydin Ah, you know, across the industry, I mean, we’re looking at a really high degree of of productivity increases. 0:13:49.640 –> 0:14:6.840 Brian Haydin Some people are saying, you know, up to 70% productivity increase and you know I’ve heard comments about like well, you know, I’m a developer and if I have, if I get 50% more work done that just means I have more work that I have to do. 0:14:6.970 –> 0:14:9.700 Brian Haydin And I think that’s kind of a misguided way of looking at it. 0:14:10.270 –> 0:14:15.970 Brian Haydin To me, I look at it as I don’t have to waste my time writing a bunch of this boilerplate stuff. 0:14:16.40 –> 0:14:28.290 Brian Haydin I can start to think about the critical problems and do a better job and investigate like you know, what are some of the edge cases that I need to consider in this particular task or this particular part of the program that I’m trying to develop. 0:14:28.460 –> 0:14:31.220 Brian Haydin So it gives you more time to actually think through problems. 0:14:31.900 –> 0:14:36.870 Brian Haydin Uh, you know, rather than writing a bunch of remedial code or, you know, just boilerplate code. 0:14:40.200 –> 0:14:47.660 Nathan Lasnoski So you wanna start by maybe just kind of talk through some of the common tasks that you’re seeing people be successful with copilot. 0:14:47.670 –> 0:14:52.320 Nathan Lasnoski This is a pretty exhaustive list, but like talk through the ones in this that you think are like really big hitters. 0:14:53.150 –> 0:14:53.520 Brian Haydin Yeah. 0:14:53.530 –> 0:14:59.960 Brian Haydin So some of this is an out there yet right, like some some this list comes from like a comprehensive list of things that you’ll be able to do. 0:15:0.560 –> 0:15:15.820 Brian Haydin So like the the pull request documentations that that’s all enterprise features that are coming out, but you know where I’ve used it in the past is I’ve got this entity framework data model and I want to write like some CRUD operations against that. 0:15:15.910 –> 0:15:28.380 Brian Haydin I can point it at that class and say, you know, help me write some, you know, read, read methods for, you know, getting an element with the specific ID but or writing my unit tests. 0:15:28.390 –> 0:15:30.520 Brian Haydin I mean, nobody likes writing unit tests. 0:15:30.530 –> 0:15:42.230 Brian Haydin I mean no developer does we all hate it, but it not only writes not only creates like good, just general coverage ones, but it can actually start to think about some of those edge cases, right? 0:15:42.240 –> 0:15:55.390 Brian Haydin So when I’m looking for values that are out of bounds and things through those problems and creates fantastic a unit tests umm and uh. 0:15:55.400 –> 0:16:6.760 Brian Haydin And also like just having it with the chat window constantly monitoring what you’re working on, you know it, it’s pretty, it can pick up on some pretty subtle things that you might miss when you’re writing code. 0:16:7.420 –> 0:16:9.710 Brian Haydin Those are probably the three biggest things that I use it for. 0:16:11.450 –> 0:16:11.840 Nathan Lasnoski Awesome. 0:16:11.850 –> 0:16:12.220 Nathan Lasnoski Awesome. 0:16:13.150 –> 0:16:19.700 Nathan Lasnoski Alright, so talk a little bit about like the IDE S and the experience and what? 0:16:19.710 –> 0:16:23.410 Nathan Lasnoski Like what practically this looks like before we go into some of the demos, Brian. 0:16:24.890 –> 0:16:26.930 Brian Haydin Well, so it is supported. 0:16:26.980 –> 0:16:34.550 Brian Haydin I mean, I use mostly Visual Studio in Visual Studio code for my development environment and it’s integrated 100% into it. 0:16:34.850 –> 0:16:37.560 Brian Haydin With the copilot, it used to be called copilot accidents. 0:16:37.570 –> 0:16:39.540 Brian Haydin Just copilot now. 0:16:39.550 –> 0:16:41.960 Brian Haydin That’s completely integrated into the IDE. 0:16:41.970 –> 0:16:49.300 Brian Haydin It’s chat window that sits on the side and it it works with the models that have been provided for it. 0:16:49.410 –> 0:17:0.760 Brian Haydin What you’re going to see in enterprise is with copilot enterprise is the ability to ground those code repos that live within your GitHub account. 0:17:1.180 –> 0:17:11.420 Brian Haydin Your GitHub organization, and it’ll become aware of these other patterns and practices and code examples so that it can provide more meaningful. 0:17:12.30 –> 0:17:21.290 Brian Haydin Uh, you know, feedback for you are more, more better examples for you ohm and again, I mean you see the jet brains up here. 0:17:21.350 –> 0:17:28.820 Brian Haydin To me, it’s like it’s it’s the best version of jet brains that I could imagine in terms of being able to give you suggestions. 0:17:28.830 –> 0:17:29.900 Brian Haydin And improvements in your code. 0:17:32.920 –> 0:17:33.390 Nathan Lasnoski Awesome. 0:17:33.560 –> 0:17:38.490 Nathan Lasnoski And you just kind of cover this a little bit, I think this is really an exciting next step. 0:17:38.500 –> 0:17:54.550 Nathan Lasnoski The whole GitHub copilot enterprise the ability for it to have more awareness of your actual code base and not be just genericized, I think is gonna be a big step forward for our ability to explainability just to like, hey, I’m going in and looking at this, what is it? 0:17:54.710 –> 0:17:55.570 Nathan Lasnoski What does this mean? 0:17:55.880 –> 0:18:4.490 Nathan Lasnoski But also being able to tailor to what how we’re developing in our organization versus maybe how someone else is developed, there’s multiple good ways to do something. 0:18:5.70 –> 0:18:8.90 Nathan Lasnoski I think that’s gonna be a really exciting next step for that as well. 0:18:9.20 –> 0:18:16.770 Brian Haydin Yeah, I would add to this too that this is a differentiator in terms of what your organizational approach is to GitHub. 0:18:16.960 –> 0:18:34.390 Brian Haydin So about seven or eight months ago, internally here at concurrency we talked about what should we really go to GitHub, you know, enterprise or should we just stick with, you know, the standard licensing and, you know, at the time we kind of like, well, there isn’t really anything in the enterprise stuff that we really need. 0:18:35.260 –> 0:18:37.40 Brian Haydin Yeah, it’s nice to get the Active Directory. 0:18:37.50 –> 0:18:38.170 Brian Haydin You’re gonna get integration. 0:18:38.670 –> 0:18:40.340 Brian Haydin Uh, maybe we get a little bit of extra. 0:18:40.380 –> 0:18:45.580 Brian Haydin Uh compute cycles out of it, but now it it’s not even a discussion. 0:18:45.650 –> 0:19:7.300 Brian Haydin I mean, if you want the the enterprise, if you want these copilot enterprise features which are gonna be hugely beneficial, I mean massively beneficial to be able to ground towards your repos, be able to have these pull request summaries and in other features that are being offered as part of this it’s a no brainer that enterprises the way to go. 0:19:10.280 –> 0:19:10.630 Nathan Lasnoski All right. 0:19:10.640 –> 0:19:12.850 Nathan Lasnoski So let’s talk about a couple examples here. 0:19:12.860 –> 0:19:14.400 Nathan Lasnoski So maybe talk through this first one, Brian. 0:19:15.500 –> 0:19:16.10 Brian Haydin Uh, yeah. 0:19:16.20 –> 0:19:19.870 Brian Haydin So I mean this is this is kind of you know the the boilerplate thing, right. 0:19:19.880 –> 0:19:21.250 Brian Haydin I wanna create a data structure. 0:19:21.260 –> 0:19:45.760 Brian Haydin I wanna create some code around this data structure that I have and maybe this is I can’t talk to this example per per se because this is not my language but, but essentially like yeah, you can just point it at a classic, you know, description of the data structure that I’m looking for and boom that just gets put out right into your window. 0:19:46.360 –> 0:19:49.650 Brian Haydin Ohm, you know natural language description. 0:19:49.700 –> 0:19:51.650 Brian Haydin I’ve got a. 0:19:51.900 –> 0:19:57.560 Brian Haydin Give me a data structure for a person and it’s gonna come up with first name, last name, all on its own. 0:19:58.730 –> 0:20:0.810 Brian Haydin You know, things like that are are really beneficial. 0:20:5.120 –> 0:20:5.290 Brian Haydin Yeah. 0:20:1.460 –> 0:20:18.20 Nathan Lasnoski I found it interesting how you could either use the chat interface like this and like describe what you want or if you just kind of like started building it in the interface it will start suggesting the next thing like name, last name, first name, last name, address. 0:20:17.930 –> 0:20:18.80 Brian Haydin Yeah. 0:20:18.30 –> 0:20:18.660 Nathan Lasnoski You know, it starts. 0:20:18.750 –> 0:20:25.170 Nathan Lasnoski It starts to kind of like anticipate what it is that you are building and suggesting the next things in that list. 0:20:25.660 –> 0:20:26.10 Brian Haydin Yeah. 0:20:26.20 –> 0:20:29.170 Brian Haydin And then I take it, I always start here and then I’ll take it to the next level. 0:20:29.180 –> 0:20:47.450 Brian Haydin Right, so now I’ve got this class here and I can continue to prompt it for service classes or create an interface that allows me to do these other things, and it has that context of the code that I’m already working with, similar to like how you’d have the context of the questions that you previously asked for. 0:20:47.580 –> 0:20:50.610 Brian Haydin And like your traditional chat GPT sessions, right? 0:20:50.680 –> 0:21:5.330 Brian Haydin So to me like it’s I can get from an idea to maybe an API exposing some of this within 15 minutes, you know versus what it might have taken me an hour or so to do it before. 0:21:6.560 –> 0:21:10.310 Nathan Lasnoski I’ll see a total like side thing like Dad moment. 0:21:19.920 –> 0:21:20.640 Brian Haydin Oh yeah. 0:21:10.370 –> 0:21:23.310 Nathan Lasnoski Like I have found that in addition to this being really useful, just for like general development, it’s great for teaching people coding like like I’ve put my son in front of this and it’s like here. 0:21:23.320 –> 0:21:29.650 Nathan Lasnoski Just describe what you want and he starts him down the road of like, he’s not having to start completely from scratch or going Googling it. 0:21:29.660 –> 0:21:31.660 Nathan Lasnoski This been awesome. 0:21:32.390 –> 0:21:39.160 Brian Haydin Yeah, in similar to that, like when you’re working with a new technology, let’s say that you wanna do some HoloLens development, right? 0:21:39.0 –> 0:21:39.200 Nathan Lasnoski Mm-hmm. 0:21:39.270 –> 0:21:47.520 Brian Haydin And you know, you’ve got some ideas about how you do it, but you don’t really wanna go through and read a bunch of documentation about how do I get started. 0:21:47.710 –> 0:22:0.20 Brian Haydin I mean, you could use it for that kind of stuff as well because there’s a ton of examples in the that are built into these models for these emerging technologies or maybe not emerging technologies, but things that you might not be familiar with. 0:22:0.610 –> 0:22:15.480 Brian Haydin So it’s not only great learning tool for somebody that’s new to programming, but it’s a it’s an an excellent learning tool for people that are trying to dive into a, you know, new technology stack or some some new components. 0:22:18.10 –> 0:22:23.390 Nathan Lasnoski Those next guys about writing unit tests talk a little bit about, like how you see this working. 0:22:24.450 –> 0:22:24.860 Brian Haydin Uh, yeah. 0:22:24.870 –> 0:22:26.60 Brian Haydin Well, I mean, I already did. 0:22:26.460 –> 0:22:26.970 Nathan Lasnoski Yeah, you did. 0:22:26.980 –> 0:22:27.300 Nathan Lasnoski That’s true. 0:22:26.110 –> 0:22:33.520 Brian Haydin Uh, you know the kind of recap it, but you know this is a good, you know, this just a good visual of how it actually works. 0:22:33.870 –> 0:22:51.660 Brian Haydin So when I write the when I use it to prompt for unit tests, I’ll usually actually get like 3 or 4 methods out of it test methods out of it so that it increases the coverage of my methods and I’m trying to test, but it’s that easy. 0:22:51.720 –> 0:22:52.570 Brian Haydin Just type it in. 0:22:52.640 –> 0:22:53.90 Brian Haydin Boom. 0:22:53.180 –> 0:22:54.490 Brian Haydin Here’s what you’re gonna test. 0:22:54.560 –> 0:22:56.290 Brian Haydin Click a button, it goes into your code. 0:22:56.300 –> 0:22:56.590 Brian Haydin You’re done. 0:22:57.670 –> 0:22:58.200 Nathan Lasnoski Ali, enough. 0:22:58.210 –> 0:23:14.120 Nathan Lasnoski I think the chat interface in and of itself has been a nice like step forward like it we’ve become really intuitive with it on other use cases and it being now part of the like development experience doesn’t you don’t have to just kind of learn how to use it in the IDE itself. 0:23:14.340 –> 0:23:17.150 Nathan Lasnoski There’s sort of sideloading this chat experience I think is really helpful. 0:23:17.850 –> 0:23:19.90 Brian Haydin 100% absolutely. 0:23:24.510 –> 0:23:31.540 Nathan Lasnoski OK, no other example of, just like, hey, I want to get kind of goes back to that never have to create a draft again. 0:23:31.610 –> 0:23:32.580 Nathan Lasnoski Like, how do I just? 0:23:32.990 –> 0:23:36.200 Nathan Lasnoski How do I do this instead of Googling it using the interface? 0:23:36.560 –> 0:23:37.660 Nathan Lasnoski How do I build this? 0:23:37.750 –> 0:23:38.660 Nathan Lasnoski Give me a prompt. 0:23:39.10 –> 0:23:42.510 Nathan Lasnoski Help me figure out how to take the next step on this particular type of thing. 0:23:43.340 –> 0:23:44.240 Nathan Lasnoski Umm yeah? 0:23:43.390 –> 0:23:51.480 Brian Haydin Yeah, this is great for developers that I’m not one of these developers that uses like get in the command line. 0:23:51.490 –> 0:24:4.640 Brian Haydin I’m not a command line developers, so shoot me, but this is a great way for command line developers who like a lot of these functions, like create a new button like your ID might have something like go hit this widget. 0:24:5.430 –> 0:24:10.960 Brian Haydin But this is a little bit more effective or efficient for developers that are really command line programmers. 0:24:15.660 –> 0:24:24.650 Nathan Lasnoski I think the idea of documenting pull requests is Rockstar, like this kind of stuff is like nobody likes documenting these. 0:24:24.660 –> 0:24:26.340 Nathan Lasnoski Nobody likes writing this kind of thing. 0:24:26.350 –> 0:24:34.610 Nathan Lasnoski Nobody even likes writing documentation in general, like the ability for it to summarize something that I did is one of the most clear and obvious. 0:24:34.620 –> 0:24:40.210 Nathan Lasnoski Even if you didn’t want to use it inline as you’re doing work just the take this off my plate. 0:24:40.220 –> 0:24:42.10 Nathan Lasnoski Part of it is fantastic. 0:24:43.540 –> 0:24:43.890 Brian Haydin Yeah. 0:24:43.900 –> 0:24:49.580 Brian Haydin And if you do this intentionally, you’re gonna see that it’s gonna allow you to think through. 0:24:49.590 –> 0:24:51.480 Brian Haydin OK, this is what it thinks I did. 0:24:53.230 –> 0:24:54.120 Brian Haydin Is that right? 0:24:54.490 –> 0:24:54.710 Nathan Lasnoski Yeah. 0:24:54.130 –> 0:25:1.580 Brian Haydin Because the is that you know the impressions that AI has probably gonna be the same impressions that somebody that’s looking at it for the first time is. 0:25:2.250 –> 0:25:3.840 Brian Haydin So did I write my? 0:25:3.880 –> 0:25:9.170 Brian Haydin Did I write this code clearly so that it’s so that’s picked up by the AI and describes it accurately? 0:25:13.300 –> 0:25:14.510 Nathan Lasnoski They might deal with commit messages. 0:25:17.70 –> 0:25:17.310 Brian Haydin Yep. 0:25:19.980 –> 0:25:21.700 Nathan Lasnoski And basic basic pattern. 0:25:23.210 –> 0:25:25.880 Nathan Lasnoski I thought this was this was really cool. 0:25:26.110 –> 0:25:30.520 Nathan Lasnoski This like idea of give me suggestions to make this better. 0:25:30.950 –> 0:25:33.820 Nathan Lasnoski I like that they’re taking it, not just the like. 0:25:33.830 –> 0:25:41.550 Nathan Lasnoski Here’s what you could do, but even like click on this to make it to replace what you previously had with what we’re suggesting. 0:25:42.650 –> 0:25:43.760 Brian Haydin Yeah, exactly. 0:25:43.770 –> 0:25:47.400 Brian Haydin And this goes back to like that whole, you know, jet brains and refactor. 0:25:48.170 –> 0:25:56.720 Brian Haydin You know, I I would use that, you know constantly and I don’t, I don’t use it anymore ever because this is way more efficient. 0:26:3.100 –> 0:26:4.430 Brian Haydin Yes, we mentioned this. 0:26:5.60 –> 0:26:6.930 Brian Haydin This is a feature that’s coming out. 0:26:6.940 –> 0:26:28.70 Brian Haydin We haven’t had a chance to play with it yet, but this is about grounding your repos in GitHub and being able to use your internal data similar like similarly to how we would ground internal documents, or how copilot 365 is gonna work. 0:26:28.280 –> 0:26:34.710 Brian Haydin This just grounds your copilot agent into your actual internal code base. 0:26:40.660 –> 0:26:48.660 Nathan Lasnoski This is a scenario that I think the sky is the limit on, and this idea of Co translation. 0:26:48.910 –> 0:26:53.20 Nathan Lasnoski We have so many customers we’re working with that they are. 0:26:53.190 –> 0:26:55.520 Nathan Lasnoski They have something old that was written in a certain language. 0:26:55.930 –> 0:26:59.640 Nathan Lasnoski They want to move to this new architecture. 0:27:0.180 –> 0:27:7.220 Nathan Lasnoski It may not be just like translate language one to language too, but we’re at the infancy of that. 0:27:7.230 –> 0:27:22.100 Nathan Lasnoski I find it really fascinating a that you can just move from one language to another and start like that, but be like I think you could start to incorporate into this other creative approaches to optimize common activities versus having to just throw people energy at. 0:27:24.590 –> 0:27:25.200 Brian Haydin Yeah. 0:27:25.250 –> 0:27:31.400 Brian Haydin I mean, I think you’re gonna see just even using the tools out of the box to be able to do this. 0:27:31.410 –> 0:27:53.470 Brian Haydin You’re gonna see like a 70% increase like imagine that you have a web forms, you know application and you wanna use a modern spot framework, you know moving forward, I mean it’s just a lot of boilerplate like look at the screen, make some stuff and and there’s some some tooling that can help you convert that. 0:27:53.660 –> 0:27:57.430 Brian Haydin But this is going to accelerate it, but you’d have to do all that work. 0:27:57.440 –> 0:28:1.420 Brian Haydin You’d have to create the new screen in Angular if you wanted to, you know, convert it to angular. 0:28:2.430 –> 0:28:8.120 Brian Haydin But you know and then and then after you did that, you’d have to think about what components am I gonna make? 0:28:8.180 –> 0:28:11.280 Brian Haydin Like how am I going to decompose this to reusable components? 0:28:11.710 –> 0:28:18.360 Brian Haydin Now all your energy is just gonna be spent on that, and you’re also gonna wind up getting tips and suggestions from copilot as well. 0:28:18.650 –> 0:28:26.190 Brian Haydin So when you’re doing a conversion or you’re trying to get rid of some end of life components, this is gonna make that job. 0:28:26.520 –> 0:28:33.610 Brian Haydin You know, I don’t know about infinitely easier because we’re trying to reduce it, so maybe infinite, you know, Infinity doesn’t apply. 0:28:33.620 –> 0:28:36.320 Brian Haydin But but it will definitely accelerate. 0:28:38.400 –> 0:28:40.500 Nathan Lasnoski Same kind of situation with code improvement, right? 0:28:40.940 –> 0:28:41.140 Brian Haydin Yeah. 0:28:45.240 –> 0:28:48.170 Nathan Lasnoski And I like one thing I like about the interface of the code. 0:29:5.920 –> 0:29:6.120 Brian Haydin Yeah. 0:28:48.180 –> 0:29:7.940 Nathan Lasnoski Improvement is a like side a side BI like how it presents it visually and I also like that you can adopt what you want like you don’t necessarily have to like take everything, but it’s almost like having a your own code review before you actually do the code review or even applying this during code reviews. 0:29:11.400 –> 0:29:12.0 Brian Haydin 100%. 0:29:13.570 –> 0:29:41.710 Nathan Lasnoski And then I think the last one I got on this, which I just thought was really kind of fun, which is like explain the code like I’m a 5 year old and because sometimes we just like we’re looking at something we really know we’re looking at like what is this, what what is the point of this thing and it trade being something that could, like I’m entering this code base for the first time, describe what this is even trying to accomplish and have it remark that back to me and have me have some perspective of what. 0:29:41.800 –> 0:29:46.750 Nathan Lasnoski Someone was actually working on here, so I stuck this at a sample program that I was working on. 0:29:46.760 –> 0:29:50.740 Nathan Lasnoski So kind of idea of explaining this to me like I’m a 5 year old. 0:29:51.690 –> 0:29:56.620 Nathan Lasnoski OK, so let’s just talk about like things to know if you’re trying to get started with this. 0:29:56.630 –> 0:30:3.560 Nathan Lasnoski Like if you’re trying to enable your development team to be AI enabled developers, what are some things that you need to know as you get started? 0:30:10.130 –> 0:30:11.820 Nathan Lasnoski So what it is you go ahead. 0:30:11.830 –> 0:30:12.590 Nathan Lasnoski Go, Brian. Sorry. 0:30:10.460 –> 0:30:17.90 Brian Haydin Ohh, it’s just gonna say, uh, we had a we had a question if this was gonna be the death of COBOL. 0:30:17.780 –> 0:30:18.280 Nathan Lasnoski Ah. 0:30:17.600 –> 0:30:26.550 Brian Haydin Uh, there’s that last I I mean, it was already dying, so maybe it’s just for all the Packer fans. 0:30:29.990 –> 0:30:30.780 Nathan Lasnoski Dagger. 0:30:26.560 –> 0:30:32.340 Brian Haydin It’s the dagger you know per say, but ohm, right? 0:30:35.780 –> 0:30:36.0 Nathan Lasnoski I know. 0:30:33.610 –> 0:30:38.840 Brian Haydin Maybe too soon after the Packers, but nonetheless what it is, what it isn’t. 0:30:39.90 –> 0:30:42.940 Brian Haydin Yeah, like I said, this is a tool that’s gonna go in my toolbox and never come out. 0:30:43.90 –> 0:30:45.260 Brian Haydin Or at least, not until something better comes around. 0:30:46.50 –> 0:30:57.140 Brian Haydin Is just going to improve your productivity in terms of, you know, getting the day-to-day, you know, kind of like Monday and tasks, you know taking care of. 0:30:57.910 –> 0:31:1.490 Brian Haydin But like again, just to call out, it’s not a replacement for developer. 0:31:1.500 –> 0:31:2.870 Brian Haydin It’s not that good. 0:31:3.180 –> 0:31:11.430 Brian Haydin You still need people that are gonna think through critically, think through the problems and but it’s gonna give you more time to do that. 0:31:11.540 –> 0:31:19.720 Brian Haydin And so I think you will see a productivity boost in terms of like can I get through like these, you know user stories and increase my teams velocity. 0:31:20.380 –> 0:31:26.170 Brian Haydin But I think the more measurable benefit is gonna be the quality of the software that you’re developing. 0:31:26.300 –> 0:31:28.0 Brian Haydin So you’re gonna see fewer bugs. 0:31:29.20 –> 0:31:38.640 Brian Haydin You know you’re going to see, you know, people thinking through the the critical problems more and so so that I think is the biggest improvement. 0:31:40.190 –> 0:31:40.520 Nathan Lasnoski Awesome. 0:31:40.530 –> 0:31:40.940 Nathan Lasnoski Thanks Brian. 0:31:43.40 –> 0:31:52.540 Nathan Lasnoski OK, this this part can be a little wonky the like getting the account stuff, especially since it’s not like just immediately integrated into your Active Directory. 0:31:52.550 –> 0:31:54.350 Nathan Lasnoski You wanna just talked about that a little second? 0:31:54.360 –> 0:31:57.710 Nathan Lasnoski Like if if someone doesn’t have it right now, like what does that process look like? 0:31:58.820 –> 0:32:0.980 Brian Haydin Yeah, so ohh. 0:32:4.980 –> 0:32:5.200 Nathan Lasnoski Yeah. 0:32:1.220 –> 0:32:6.290 Brian Haydin I mean, we probably need about 20 minutes to talk to you about the entire process is but. 0:32:6.780 –> 0:32:16.870 Brian Haydin But yeah, so you’re going to need a GitHub account and a GitHub organization that’s set up in order for you to enable copilot for your organization. 0:32:16.880 –> 0:32:33.370 Brian Haydin So if you’re using DevOps today that you know you’re probably you’re going to need to start using GitHub, so you’re going to want to set up an organization, and then there’s a decision to be made about whether or not you want to take advantage of the enterprise. 0:32:33.820 –> 0:32:54.970 Brian Haydin You’re gonna have a choice between basic licensing and enterprise licensing, and you know, I would strongly encourage everybody to consider the enterprise licensing because it’ll give you the ability to ground your copilot instance, to your your internal repos, most development. 0:32:56.930 –> 0:33:11.580 Brian Haydin Most development work is using Git repositories anyways, so you can still continue to use ADO, but in terms of how a housing your repositories you’re gonna wanna move that to to GitHub. 0:33:11.630 –> 0:33:12.500 Brian Haydin You know moving forward. 0:33:15.400 –> 0:33:15.810 Nathan Lasnoski Awesome. 0:33:15.860 –> 0:33:23.390 Nathan Lasnoski We talked about IDs already this a little bit of a comparison of like the per user cost per month. 0:33:23.400 –> 0:33:27.30 Nathan Lasnoski So enterprise is certainly more expensive, but certainly worth it. 0:33:31.100 –> 0:33:31.260 Brian Haydin Yeah. 0:33:27.40 –> 0:33:32.110 Nathan Lasnoski Just the ability to ground is a pretty huge embankment, so that’s on the wait list. 0:33:32.120 –> 0:33:32.940 Nathan Lasnoski It’s coming soon though. 0:33:32.950 –> 0:33:35.570 Nathan Lasnoski I mean talk in February, which is like days away. 0:33:36.320 –> 0:33:36.540 Brian Haydin Yep. 0:33:39.210 –> 0:33:44.320 Nathan Lasnoski There is an interesting uh button in copilot. 0:33:44.330 –> 0:33:47.170 Nathan Lasnoski Do you wanna just talk about like the copyrighted material thing? 0:33:47.180 –> 0:33:48.290 Nathan Lasnoski That’s cause is a common concern. 0:33:49.270 –> 0:33:50.320 Brian Haydin Yeah, it is. 0:33:50.430 –> 0:34:0.610 Brian Haydin It’s literally just a button, but so if your organization has policies around using open source code or copyrighted material. 0:34:2.150 –> 0:34:16.20 Brian Haydin GitHub copilot has taken care of that problem, addressing concerns, so it will look for any of the codes suggestions that it’s made that would match public sources and potentially be in violation. 0:34:16.660 –> 0:34:23.300 Brian Haydin And so this is part of your organization settings and would be applied to everybody that was using it. 0:34:23.980 –> 0:34:31.140 Brian Haydin So if you if you feel like this is something you don’t wanna do because of the potential copyright risks, this is a way that you could avert that. 0:34:31.970 –> 0:34:33.920 Nathan Lasnoski What do you think it’s funny about this one? 0:34:33.990 –> 0:34:34.740 Nathan Lasnoski Is this like? 0:34:35.890 –> 0:34:38.470 Nathan Lasnoski I mean it’s it’s it’s it’s an advancement, right? 0:34:38.940 –> 0:34:39.100 Brian Haydin Yeah. 0:34:38.830 –> 0:34:41.120 Nathan Lasnoski But everybody does this already anyway. 0:34:41.230 –> 0:34:44.630 Nathan Lasnoski Like they’re already copying and pasting stuff from, like public repos. 0:34:46.760 –> 0:34:47.530 Nathan Lasnoski But it’s it’s. 0:34:47.540 –> 0:34:51.660 Nathan Lasnoski I think it’s neat that it it’s almost like building that out of my environment in a sense. 0:34:52.130 –> 0:34:52.320 Brian Haydin Yep. 0:34:53.970 –> 0:34:54.520 Nathan Lasnoski OK. 0:34:54.730 –> 0:34:56.920 Nathan Lasnoski So we’re gonna transition the conversation now. 0:34:56.930 –> 0:35:5.500 Nathan Lasnoski We just talked a lot about like using AI to be an enabled developer in the sort of high code sense we’re gonna move into like the semi code space, the. 0:35:7.250 –> 0:35:12.0 Nathan Lasnoski More like self service development area. 0:35:12.110 –> 0:35:18.330 Nathan Lasnoski Talking about Copilot studio in copilot studio, you know if you think about the previous graphic that we had. 0:35:18.340 –> 0:35:29.120 Nathan Lasnoski This is like your full code developer track what’s happening in the sort of low code, no code, semi code sort of space is a JSON to that. 0:35:29.190 –> 0:35:51.450 Nathan Lasnoski You’re seeing similar advancements if you have listened to any talks that Charles La Rama has given the Steve P and Microsoft, who owns power apps, he is has a has a goal, is like North Star is to make developing applications as easy as building a PowerPoint like that. 0:35:51.460 –> 0:35:54.610 Nathan Lasnoski PowerPoints right now are just like something anybody can do, right? 0:35:54.720 –> 0:36:4.910 Nathan Lasnoski He wants building applications and presenting data to be something that is as easy for an individual to do as building a PowerPoint presentation. 0:36:5.160 –> 0:36:9.130 Nathan Lasnoski And it’s not quite there yet, but you can see how they’re making progress. 0:36:9.350 –> 0:36:21.930 Nathan Lasnoski So simple power apps development we’ve been seeing that happen a long time or any kind of low code, no code sort of space like there’s these sort of middle tier of applications that we’re seeing advance in their capability. 0:36:21.940 –> 0:36:50.570 Nathan Lasnoski We’re seeing even full scale development capabilities like ML OPS and DevOps start to be part of that picture, but which are also starting to see is that English language centric development or even translated language centric development is starting to enter into how people are building applications even in the sense that we used it in the GitHub copilot space, powerapps and other capabilities in copilot studio are starting to do the same thing. 0:36:50.580 –> 0:37:7.80 Nathan Lasnoski So what we’re going to talk about is especially copilot studio, and we’ll just kind of intro a little bit of what they’re even doing in power apps to achieve efficiency for the whole community that wants to build applications that are in that sort of middle zone or even higher as they start to become very capable. 0:37:8.120 –> 0:37:21.630 Nathan Lasnoski So one thing that I think is really cool, you can go to make dot powerapps right now to try this out which is you can describe the app you want it to build and it will start that application for you. 0:37:21.700 –> 0:37:36.570 Nathan Lasnoski Now there’s a whole lot it doesn’t do like I think it’s very much in its infancy, but it’s very exciting just to think about the idea of what would it mean if I individuals were being able to build these themselves, to track their ideas or take action on something they think is possible. 0:37:36.720 –> 0:37:44.180 Nathan Lasnoski So I just went out there and said I wanna track an app that will track AI ideals because we do this all the time, like we’re meeting with customers, executive teams. 0:37:44.190 –> 0:37:50.560 Nathan Lasnoski We’re building these lists of AI ideas, and then we’re articulating that back out, we want to build an app for it. 0:37:50.760 –> 0:37:52.250 Nathan Lasnoski So let’s build a power app. 0:37:52.260 –> 0:37:58.270 Nathan Lasnoski So the app could capture the name, category, description, revenue, production, operational savings priority and dependencies. 0:37:58.780 –> 0:38:3.690 Nathan Lasnoski I can put that in a spreadsheet, but historically, like most people couldn’t put that into an application. 0:38:3.940 –> 0:38:8.600 Nathan Lasnoski Now with power apps you can so it immediately takes that builds. 0:38:8.610 –> 0:38:25.120 Nathan Lasnoski This table builds the interface that will exist for it to be able to function as an application, and then as one thing I thought was really neat about it is side loaded into the experience are this ability to edit as I went. 0:38:25.130 –> 0:38:30.540 Nathan Lasnoski So for example, I said add a column for tracking if it produces new revenue and how much per year. 0:38:30.850 –> 0:38:37.260 Nathan Lasnoski But then I notice like, oh, shoot, this wasn’t on the screen like there’s a slider here and it was over hidden on the other side of the screen. 0:38:37.270 –> 0:38:40.800 Nathan Lasnoski So then I had two tables like 2 columns for that same thing. 0:38:40.970 –> 0:38:42.380 Nathan Lasnoski So I said undo it. 0:38:42.470 –> 0:38:45.780 Nathan Lasnoski Undo the change and it removed it because I hadn’t you put any data there yet. 0:38:45.870 –> 0:38:52.840 Nathan Lasnoski I find the interface to be very intuitive and something that, just like any general individual, appeal to take advantage of. 0:38:52.880 –> 0:39:3.950 Nathan Lasnoski So I think it’s really exciting where the ability to create applications is going and how we can drop that down to other levels of our organizations to enable them to be makers as well. 0:39:4.330 –> 0:39:8.170 Nathan Lasnoski And this is happening as well inside of the power automate space. 0:39:8.340 –> 0:39:23.60 Nathan Lasnoski So the ability to describe what you want and have it build the workflow that I intentionally am interested in have it make the connections and even make adjustments to those based upon the chat experience on the right hand side. 0:39:23.180 –> 0:39:45.860 Nathan Lasnoski So essentially walking an individual through being a developer, but not having to put the the only way you get to develop things is as you learn to code like no, I think what’s gonna happen is more and more often the creative skill is what’s going to really be accentuated within us, whether you know how to write that code in full code or not. 0:39:45.980 –> 0:39:50.230 Nathan Lasnoski And this is going to enable everyone to be able to participate within the development process. 0:39:52.70 –> 0:39:52.600 Nathan Lasnoski And now? 0:39:52.660 –> 0:39:55.410 Nathan Lasnoski Now what we’re gonna do is Ajay. 0:39:55.420 –> 0:39:59.940 Nathan Lasnoski You’re gonna kind of start talking through a little bit about copilot studio and this is really exciting. 0:40:0.20 –> 0:40:1.630 Nathan Lasnoski Like what they have done? 0:40:1.640 –> 0:40:11.870 Nathan Lasnoski I don’t know if anybody who played with power virtual agents before, but like the difference between power virtual agents in which you can now do in copilot studio is pretty stunning. 0:40:12.200 –> 0:40:13.570 Nathan Lasnoski So Ajay, take it away. 0:40:13.580 –> 0:40:18.440 Nathan Lasnoski I’d love for you to talk through this and HSO gonna demo a couple examples of what you can do. 0:40:20.440 –> 0:40:21.10 Ajay Ravi Thanks, Nate. 0:40:21.470 –> 0:40:24.960 Ajay Ravi So let me just share my screen real quick. 0:40:34.120 –> 0:40:34.640 Ajay Ravi OK. 0:40:34.880 –> 0:40:44.390 Ajay Ravi So we talked a little bit about like the app and the automation side of low code and now let’s take a look at what is gopa let’s studio. 0:40:44.640 –> 0:40:59.490 Ajay Ravi So copilot studio was something which Microsoft introduced as part of last year’s Ignite, and it is a low code platform that allows you to build intelligent chat bots using either generative AI or even your own custom data source. 0:40:59.680 –> 0:41:11.680 Ajay Ravi So it was something which Microsoft introduced last year and it’s kind of an extension to power virtual agent with lot of additional capabilities like generative AI and all those things. 0:41:11.690 –> 0:41:18.60 Ajay Ravi So, but you do have the ability to kind of enable or disable those functionalities as well. 0:41:18.70 –> 0:41:22.580 Ajay Ravi If you would like, but it gives a lot of times, so capabilities for that. 0:41:22.750 –> 0:41:29.960 Ajay Ravi So on the right hand side you can see an example of a screenshot of how the copilot studio looked like. 0:41:30.70 –> 0:41:43.70 Ajay Ravi So you don’t have to like write any additional code in order to build any of these interfaces, and you can purely or build the logic based on the graphical interface and then kind of like navigate through the different topics. 0:41:43.320 –> 0:41:55.410 Ajay Ravi So you can kind of like connect your chat GPT TTS or like you know generative AI or if you would like to build any custom logic with just the graphical interface without any code. 0:41:55.520 –> 0:41:57.250 Ajay Ravi So you can definitely do that as well. 0:41:57.260 –> 0:42:7.880 Ajay Ravi And then if you want to like connect to, it’s just like uh and SAP dynamics 365 sales four or any of those data if you want to like bring that to the OR chat bot. 0:42:7.890 –> 0:42:10.60 Ajay Ravi So you do have the ability to do that as well. 0:42:10.70 –> 0:42:16.640 Ajay Ravi So there are over 1100 plus prebuilt plugins available in copilot studio. 0:42:17.130 –> 0:42:18.890 Ajay Ravi So you can use any one of those. 0:42:18.900 –> 0:42:25.700 Ajay Ravi You can also use power automate to kind of like bring data from an external source back to copilot studio. 0:42:25.710 –> 0:42:30.450 Ajay Ravi So you can either do that, or if you want to like use advance or development capabilities. 0:42:30.460 –> 0:42:52.240 Ajay Ravi You can also build ohm, a custom plugin and then connect it back to the copilot schedule so that way either you can bring or data from your own customized data source, or you can additionally use generative AI functionality to kind of like look at data from public websites or internal websites like SharePoint and Things like that. 0:42:52.530 –> 0:42:58.280 Ajay Ravi So it gives a lot of additional capabilities of which virtual agent did not have previously. 0:42:58.390 –> 0:43:6.450 Ajay Ravi So it’s a really cool thing where users can kind of like interact on a daily basis, make their day-to-day life a bit more harm predictive. 0:43:10.570 –> 0:43:15.220 Ajay Ravi So going into a little bit about the architecture side of things. 0:43:15.510 –> 0:43:20.590 Ajay Ravi So copilot Studio definitely comes with other authentication side of things. 0:43:20.600 –> 0:43:31.440 Ajay Ravi So you can use Active Directory or let’s single sign on functionality as well in order to authenticate and make sure that the bot is communicating with users who have write access. 0:43:31.810 –> 0:43:49.210 Ajay Ravi So that is definitely a functionality with copilot studio and in addition to that you do have the ability to add additional policies like for example or there is data loss prevention policies that you can set up at the tenant level and the environmental level. 0:43:49.220 –> 0:43:56.230 Ajay Ravi So that way you can or restrict or allow certain authentication, certain connectors and all those things. 0:43:56.380 –> 0:44:12.110 Ajay Ravi So the copilot studio is Bell on top of data loss, and Microsoft Power platform, so it allows you to kind of like make sure that only the right of authentication or if you want to disable HTTP connectors you can do that. 0:44:12.120 –> 0:44:23.980 Ajay Ravi So you do have the additional ability to allow and disable all those connectors, so it comes with a lot of additional data policies as well. 0:44:24.930 –> 0:44:29.750 Ajay Ravi And then if it take a look at how some of how the conversation works. 0:44:29.760 –> 0:44:40.780 Ajay Ravi So it is built on top of Azure bot service and then whenever you set up a port it already comes with some system or generated entities. 0:44:40.790 –> 0:44:48.130 Ajay Ravi So entities are basically how the bot understand what the user is talking about, like for example is color ohm. 0:44:48.170 –> 0:44:49.360 Ajay Ravi You know all those numbers. 0:44:49.370 –> 0:44:54.640 Ajay Ravi And so that really helps the bot understand what the user is talking about. 0:44:54.760 –> 0:45:3.700 Ajay Ravi So by default, whenever you set up a port, it already comes with a bunch of system ohh system generated entities. 0:45:3.950 –> 0:45:17.540 Ajay Ravi But in addition to that, if you would like to add additional logics where you want to bring in your business or entities or like business specific conditions or like logic. So you do have the ability to create or the custom as well. 0:45:17.710 –> 0:45:20.340 Ajay Ravi So that way or the bot will kind of like understand. 0:45:20.350 –> 0:45:33.290 Ajay Ravi How about the user is asking about and then accordingly navigate the conversation and uh, if you look at uh, the generative AI side of things, so this is a feature which was not available with power virtual agent. 0:45:33.300 –> 0:45:35.10 Ajay Ravi And with this copilot studio. 0:45:35.140 –> 0:45:45.110 Ajay Ravi So you do have the ability to enable that functionality, but if you do not want to like use generative functionality, so you have the ability to disable that as well. 0:45:45.120 –> 0:45:54.770 Ajay Ravi So if you enable, you can add up to four customs or it’s or like you know, if you want to like add internal sites like SharePoint and OneDrive. 0:45:54.780 –> 0:46:0.650 Ajay Ravi So you do have the ability to do that as well, but if you want to like, uh, disable that, you can totally do that. 0:46:0.660 –> 0:46:7.930 Ajay Ravi And then create custom logic using just graphical interface without any code and then or navigate through the conversation. 0:46:8.180 –> 0:46:11.700 Ajay Ravi So it comes with a lot of additional logic. 0:46:11.710 –> 0:46:21.290 Ajay Ravi Has Phil and in addition to that you can trigger various or connectors like all 1100 plus prebuild all connectors that are already available. 0:46:21.460 –> 0:46:26.320 Ajay Ravi So you can either use those or create custom or connectors or you know you can. 0:46:26.330 –> 0:46:41.890 Ajay Ravi You can also use power automate in order to fetch data from an external system like dynamics 365 or you know SAP and all those things and at any point of time if you want to escalate or to an actual agent. 0:46:41.940 –> 0:46:51.530 Ajay Ravi So you do have the ability to do that as well, so you can escalate it to a different application like dynamics 365, customer service or service now or any of those tools. 0:46:51.540 –> 0:46:59.870 Ajay Ravi So you do have the ability to escalate and the agent will have the live chat or whatever the user have typed so far. 0:47:0.60 –> 0:47:2.630 Ajay Ravi So they will be able to get that information as well. 0:47:2.900 –> 0:47:21.60 Ajay Ravi So it gives a lot of abilities and then you can kind of like connect the copilot studio or to different places like for example or you can connect it in Microsoft Teams or so that it is easy to or easy for the users to kind of like get that information and then share on a daily basis. 0:47:21.210 –> 0:47:43.590 Ajay Ravi So you do have all the abilities to connect to different platforms as well, and also since everything is built on top of data, worse and our power platforms, so it makes it really easy for rapid development of the bots and then kind of like deploy it to higher environment and then production use and then you can do in an iterative approach to improve the performance as well. 0:47:43.940 –> 0:47:50.130 Ajay Ravi So uh, it really makes all the board of development and then deployment relatively easy. 0:47:54.610 –> 0:48:3.510 Ajay Ravi So if you look at a little bit about why copilot studio so usability is one of the great advantage of using our copilot studio. 0:48:3.680 –> 0:48:9.310 Ajay Ravi So you are basically using literally no code and then you using graphical builder. 0:48:9.440 –> 0:48:20.110 Ajay Ravi You can create all the custom logics and everything so that is one of the biggest advantage and you do not actually require developer already data scientists to build all the open AI. 0:48:20.240 –> 0:48:21.990 Ajay Ravi So all these things are done natively. 0:48:22.530 –> 0:48:24.140 Ajay Ravi Uh, in the copilot studio? 0:48:24.310 –> 0:48:31.140 Ajay Ravi And it has the ability to understand the natural language capabilities and then it can kind of like reuse contents. 0:48:31.150 –> 0:48:36.400 Ajay Ravi So if you created one custom logic you can reuse throughout different context as well. 0:48:36.530 –> 0:48:38.960 Ajay Ravi So that makes the life of the developers. 0:48:38.970 –> 0:48:41.190 Ajay Ravi Or like who’s building the bot ACC as well? 0:48:42.780 –> 0:48:57.630 Ajay Ravi And if you look at the productivity side of things, you can, since it is Bill on top of power platform, so you can easily improve the bot performance by looking at the analytics feature that is available in our copilot studio. 0:48:57.740 –> 0:49:7.50 Ajay Ravi So you can, you know, look at the performance, analyze and then make iterative approaches to improve and increase the performance and add additional functionality. 0:49:7.300 –> 0:49:13.810 Ajay Ravi And you can also add that to Microsoft Teams, so that makes it is pretty ohh easy to use as well. 0:49:15.650 –> 0:49:25.900 Ajay Ravi Ohh and if you look at all the extensibility side of things, so even though, uh, copilot studio is a low code tool so it doesn’t mean that you cannot use our pro code. 0:49:25.910 –> 0:49:36.320 Ajay Ravi So you definitely have the ability to create custom or plugins and you can connect that to your data source and the chat bot should be able to retrieve that information. 0:49:36.490 –> 0:49:56.230 Ajay Ravi So all those are additional pro dev extensibility feature can also be used in copilot studio and it really comes with 1100 plus or you know connectors and it is easy to kind of like integrate with other power platform tools like our power app or Power pages and all those things. 0:49:56.240 –> 0:50:1.360 Ajay Ravi So it’s pretty easy to kind of like our connect and do those things. 0:50:2.530 –> 0:50:20.600 Ajay Ravi And finally, if you look at the scalability side of things, so you can easily reuse some of the content throughout the bar and then all you can look at the analytics side of things to improve the performance and then kind of like doing an iterative approach to deploy to higher environment. 0:50:20.810 –> 0:50:26.530 Ajay Ravi So these are some of the reasons why copilot Studio might be beneficial for your organization. 0:50:30.730 –> 0:50:37.190 Ajay Ravi Ohh, let’s take a quick look at how the generative answers consideration and how that works. 0:50:37.390 –> 0:50:43.740 Ajay Ravi So like I mentioned, generative AI functionality is something which you can enable or disable for the bot. 0:50:43.810 –> 0:50:55.840 Ajay Ravi So if you disable the generative AI functionality, then in that scenario you can create custom logic and that would kind of like navigate the conversation or throughout the port. 0:50:55.930 –> 0:51:3.530 Ajay Ravi But if you enable the generative AI functionality so you can, uh connect to public data like all public website. 0:51:3.540 –> 0:51:6.490 Ajay Ravi So you can add to your chat bot. 0:51:6.500 –> 0:51:17.550 Ajay Ravi So currently all the limit to add is up to four URL for the public website and then in case of our SharePoint so you can add up to four SharePoint URLs. 0:51:17.600 –> 0:51:38.540 Ajay Ravi So there is this kind of like the condition or like you know limit or currently that is available and then you can go up to two or subpages as well of currently and in addition to that you do have the ability to connect to our our Azure open AI on your data as well or in order to generate or generative AI response. 0:51:38.610 –> 0:51:43.700 Ajay Ravi So you can definitely do that as well as a part of generative AI functionality. 0:51:46.190 –> 0:52:4.600 Ajay Ravi In addition to that, if you want to like upload certain text related documents like Word document and all those things, so you can upload that as well to that chat GPT or like you know or copilot studio and then the bot will be able to kind of like or retrieve information from that uploaded document. 0:52:4.750 –> 0:52:14.320 Ajay Ravi So that is again another way to kind of like retrieve or generate generative AI response and finally you can create custom data. 0:52:14.380 –> 0:52:20.740 Ajay Ravi So using GITP request or power automate, you can kind of like get that information and bring it back to. 0:52:22.340 –> 0:52:22.970 Ajay Ravi Generac. 0:52:23.340 –> 0:52:25.30 Ajay Ravi Studio to generate the response. 0:52:25.160 –> 0:52:42.10 Ajay Ravi So this way you can kind of like retrieve all those information from various sources and then you can Azure open AI would kind of like do the summarization side of things and then response validation happen and then it is returned to the copilot studio as response. 0:52:46.110 –> 0:52:58.250 Ajay Ravi And finally, if you look at, uh, some of the channel integration and how they handle works, so you can easily integrate the copilot studio or in Microsoft Teams or like in a couple custom websites. 0:52:58.260 –> 0:53:2.280 Ajay Ravi So you can definitely do that and like in Facebook mobile app. 0:53:2.290 –> 0:53:11.800 Ajay Ravi So all those things are compatible and you can kind of like extend and add the copilot studio port to the all these different platforms. 0:53:12.190 –> 0:53:23.20 Ajay Ravi And at any point of time, if you want to kind of like seamlessly transferred that to a live agent with all the context, so you can definitely do that as well. 0:53:23.290 –> 0:53:33.30 Ajay Ravi So you can give it give the conversation or move the conversation to dynamics 365 customer service service now or like you know all these different engagement hubs as well. 0:53:33.170 –> 0:53:40.930 Ajay Ravi So this will really help the agent have the context of what the user was talking about and then accordingly hand it over. 0:53:45.200 –> 0:53:52.390 Ajay Ravi So a little bit about some of the use cases and how some of these things can be used within the organization. 0:53:52.680 –> 0:53:58.570 Ajay Ravi So in this scenario I have kind of like added two different sites. 0:53:58.580 –> 0:54:9.650 Ajay Ravi One is for an internal SharePoint site and in addition to that I’ve also added a public website and using generative AI or it will kind of like kind of get the response. 0:54:9.700 –> 0:54:32.40 Ajay Ravi So before I go into the live demo or just kind of like showing so it’s kind of like asked how do I reset my password or it looked at the internal SharePoint site and returned with other URL’s that would directly take you to that appropriate page and then in the scenario or you can see that I asked like if concurrency help with or data and AI projects. 0:54:32.50 –> 0:54:38.60 Ajay Ravi So it looked at the concurrency public website and or the results along with all the URL. 0:54:38.790 –> 0:54:42.240 Ajay Ravi So this way you can for the same chat bot. 0:54:42.250 –> 0:54:47.590 Ajay Ravi You can kind of like connect with public side as well as our internal sites as well. 0:54:50.480 –> 0:55:1.710 Ajay Ravi And another example is, so you can also create or like read records from a different system using using copilot studio and power automate. 0:55:1.760 –> 0:55:12.290 Ajay Ravi So in this scenario I’m creating an order record in Dynamics 365 from chat from the copilot studio by just entering couple of information. 0:55:12.500 –> 0:55:17.770 Ajay Ravi So this way users can kind of like without having to go to multiple system. 0:55:17.780 –> 0:55:33.890 Ajay Ravi You can directly go ahead and create records multiple systems and finally another thing would be you know you can easily integrate to our Microsoft Teams and that would allow you to kind of like, you know, chat directly from teams. 0:55:37.170 –> 0:55:57.780 Ajay Ravi It quick look at how the portal looks like, so copilot is the portal which you can use for using the copilot studio and you can see that I’ve created a port or test port concurrency board and in here you can see if you go to the generative AI side of things. 0:55:57.830 –> 0:56:14.610 Ajay Ravi So you can see I have added few public website as well as one internal website and so this is the place where you can kind of like test all those functionality and you do have the ability to kind of like connect this to teams or different channels as well. 0:56:14.980 –> 0:56:18.650 Ajay Ravi So let’s take a look at how it works. 0:56:18.660 –> 0:56:20.920 Ajay Ravi So I’m just going to type. 0:56:21.40 –> 0:56:21.720 Ajay Ravi Who is? 0:56:26.410 –> 0:56:41.460 Ajay Ravi OK, So what it does is so since we have the website public website added so it pretty much went to ohh that location and then it returned or what concurrency does and then some results as well. 0:56:41.660 –> 0:56:50.400 Ajay Ravi So if I go to all the URL so you can see that even though I added just, it went to about us and then return that results. 0:56:50.410 –> 0:57:3.490 Ajay Ravi So it’s a good way to kind of like see how that conversation went and you can see that these are some of the connectors that got triggered automatically and then it went ahead and returned those results. 0:57:5.100 –> 0:57:12.990 Ajay Ravi Another thing which you can argue check is so I’m going to check something regarding internal or SharePoint site. 0:57:13.0 –> 0:57:17.560 Ajay Ravi So I’m going to check how to reset my password. 0:57:19.450 –> 0:57:27.360 Ajay Ravi So right now we are going to look at the internal SharePoint side in order to resolve or get the result. 0:57:27.450 –> 0:57:30.300 Ajay Ravi So you can see that it returned auto reset your password. 0:57:30.310 –> 0:57:34.60 Ajay Ravi You can go to password reset page and answer a few security questions. 0:57:34.350 –> 0:57:48.250 Ajay Ravi So if I go to the URL so you can see that how this equation or like with answers for FAQ or which says like how do I reset the password so it pretty much looked at how the SharePoint location and then returned that. 0:57:48.400 –> 0:57:49.190 Ajay Ravi Ohh just said. 0:57:51.580 –> 0:58:0.470 Ajay Ravi And in addition to that, so like I mentioned, you can also add documents to the copilot studio that will allow you to or get that result. 0:58:0.480 –> 0:58:8.490 Ajay Ravi So if I go ahead and ask something like this, there presence for. 0:58:15.680 –> 0:58:25.110 Ajay Ravi So I’m just checking if there’s a process for emergency leave, so all these information are attached to a document Word document. 0:58:25.120 –> 0:58:32.290 Ajay Ravi So you can see that it returned the Word document URL and when you click that it will even download to your system. 0:58:32.300 –> 0:58:39.350 Ajay Ravi So this way you can easily access so you can see that in the document section I’ve added like few documents. 0:58:39.360 –> 0:58:44.540 Ajay Ravi So this way I’ll copilot studio or allow us to kind of like get that information and return back. 0:58:47.340 –> 0:58:59.60 Ajay Ravi And and this scenario which I just want to quickly show is how you can trigger power automate to kind of like ohh talk with a different system and then return results. 0:58:59.70 –> 0:59:6.210 Ajay Ravi So I’m just going to type create purchase order so it pretty much asks like what’s my email address. 0:59:6.220 –> 0:59:10.620 Ajay Ravi So I’m just going to type my email address. 0:59:12.810 –> 0:59:15.220 Ajay Ravi So now it has like which device? 0:59:15.290 –> 0:59:17.820 Ajay Ravi Uh, would I like to place order for? 0:59:17.830 –> 0:59:30.720 Ajay Ravi So if I type all laptops so it will go ahead and create a record in a different system and it will return back with the order ID and if I go to dynamics 365. 0:59:30.770 –> 0:59:34.260 Ajay Ravi So now you can see that in your record got created just now. 0:59:34.470 –> 0:59:47.140 Ajay Ravi So this way now from the chat experience you can directly interact with multiple systems and then I’ll get the results or or you can even use this for record creation or updation and all those things. 0:59:52.20 –> 0:59:52.470 Nathan Lasnoski Awesome. 0:59:52.770 –> 0:59:53.620 Nathan Lasnoski Thank you, Ajay. 0:59:52.150 –> 0:59:54.260 Ajay Ravi So. Umm. 0:59:53.630 –> 0:59:54.900 Nathan Lasnoski That’s that was fantastic. 0:59:56.160 –> 1:0:1.60 Nathan Lasnoski OK, so we want to make sure that we wrap this up. 1:0:1.190 –> 1:0:8.20 Nathan Lasnoski The two things I wanna make sure I show you this is like a total handwritten document, but I think it’s it helps to understand how this all fits together. 1:0:8.900 –> 1:0:18.700 Nathan Lasnoski So many of you were starting to take advantage of M365 copilot M365, copilot can sit inside of this picture and answer questions. 1:0:18.970 –> 1:0:31.600 Nathan Lasnoski You could also have an individual interacting with a copilot studio based application directly from within teams, or it can plug into the M365 copilot experience. 1:0:31.770 –> 1:0:42.520 Nathan Lasnoski So it can function as a as a plugin between these two agents, so that can fit together and then some of our companies are working with. 1:0:42.530 –> 1:0:55.660 Nathan Lasnoski They’ve also spinning up like custom GPTS that are akin to copilot studio and or maybe it’s a specific use case that it answers specific kinds of things really well, and that also can fit into the entire picture. 1:0:55.670 –> 1:1:17.150 Nathan Lasnoski So all these sort of fit inside of the same space and eventually as we kind of walked forward here, we’re gonna find is that copilot, copilot, studio apps, full code, ML apps built inside of Azure AI studio to do certain custom GPTS are all going to interact with each other through the vehicles of plugins. 1:1:17.280 –> 1:1:20.30 Nathan Lasnoski And sometimes you might be hitting them directly from teams. 1:1:20.160 –> 1:1:27.730 Nathan Lasnoski Sometimes you might be hitting them, copilot, but what you’re trying to create is an ecosystem where all of this is in the same sandbox. 1:1:27.820 –> 1:1:39.520 Nathan Lasnoski You’re trying to have all these talk together in the same space and then good sense that architecture exists for that to happen, but it needs to also be intentional, that like people aren’t just like creating all their own things. 1:1:39.670 –> 1:1:42.280 Nathan Lasnoski There’s a way for this all to be brought together. 1:1:42.510 –> 1:1:45.950 Nathan Lasnoski So we’ve talked about how it makes the difference. 1:1:46.30 –> 1:1:47.380 Nathan Lasnoski How do you get going with this? 1:1:47.390 –> 1:1:52.20 Nathan Lasnoski There’s a couple ways that I’d like to suggest, and they’re in your sort of call to action form. 1:1:52.240 –> 1:2:3.610 Nathan Lasnoski The first is we would love to sit with your team to help you get started with GitHub copilot with copilot studio to help this become an asset to you to be a partner to your dev organization. 1:2:3.620 –> 1:2:4.960 Nathan Lasnoski This is something we do quite a bit. 1:2:4.970 –> 1:2:6.720 Nathan Lasnoski We’re not a staff log business. 1:2:6.980 –> 1:2:17.340 Nathan Lasnoski We bring high scale capabilities to customers to help them do accelerate their own teams and this is an opportunity for for us to be a partner to you in that space to envision this. 1:2:17.350 –> 1:2:24.200 Nathan Lasnoski And we’ll invest in the envisioning to help you think about how can you use copilot studio, where are the business opportunities? 1:2:24.290 –> 1:2:30.840 Nathan Lasnoski How can I use copilot in GitHub GitHub to be able to accelerate my development team in a very practical way? 1:2:30.930 –> 1:2:37.160 Nathan Lasnoski In experiment where that can be path forward, we also can help you evaluate and take action on that scenario. 1:2:37.490 –> 1:2:41.720 Nathan Lasnoski Some cases we’re seeing companies say how can you accelerate moving from this thing? 1:2:41.730 –> 1:3:2.40 Nathan Lasnoski I have to this thing I have or how can you accelerate these types of activities of my particular developers to help them be able to use this as a normative way of doing work within your organization and then ultimately your goal is to move this into a scaled production norm as the teams become very, very comfortable with it. 1:3:2.110 –> 1:3:4.160 Nathan Lasnoski So our goal is to help you get there. 1:3:4.250 –> 1:3:8.230 Nathan Lasnoski So our call to action is what we love to start with is just that group and visioning session. 1:3:8.240 –> 1:3:14.140 Nathan Lasnoski That’s us investing in you to learn more about your process and your organization, and we’d love to take that step. 1:3:14.150 –> 1:3:18.570 Nathan Lasnoski So as you close down today, please take the time to fill out the survey. 1:3:18.580 –> 1:3:23.480 Nathan Lasnoski We just want to know how you like the session and B we love to know what we can do better and see. 1:3:23.490 –> 1:3:24.200 Nathan Lasnoski We wanna help you. 1:3:24.210 –> 1:3:32.410 Nathan Lasnoski So if you if be worthwhile US spending some time together learning about your organization, please indicate that in the form and we’d love to take that action. 1:3:32.480 –> 1:3:34.570 Nathan Lasnoski So I hope everyone has a great rest of your day. 1:3:34.650 –> 1:3:41.900 Nathan Lasnoski Thank you for going a little bit long and thank you especially to Brian and Ajay for the activities that you walk through today. 1:3:43.800 –> 1:3:44.380 Brian Haydin Thanks everybody. 1:3:47.330 –> 1:3:47.590 Ajay Ravi Girl.