View Recording: Mastering Microsoft Copilot for Peak Productivity: New Features & Essential Strategies

Corey Milliman So today we’re going to be going over some new copilot features that came out with Wave 2. 0:0:29.48 –> 0:0:33.248 Corey Milliman So to get started today, we’re going to talk about copilot pages. 0:0:33.248 –> 0:0:37.448 Corey Milliman We’re going to be talking about copilot and Python And Excel. 0:0:38.348 –> 0:0:44.868 Corey Milliman And then we’re going to be going over how to create copilot agents to do some of the things that you don’t wanna do every day. 0:0:45.548 –> 0:0:47.868 Corey Milliman So first of all copilot pages right? 0:0:48.148 –> 0:0:57.868 Corey Milliman So copilot pages are interactive workplaces within loop and that allows us to Co create and edit content together in real time. 0:0:58.348 –> 0:1:6.388 Corey Milliman And what that does is when we are working with copilot to come up with ideas and generate content, we can share this in real time with other people. 0:1:7.168 –> 0:1:7.608 Corey Milliman So that. 0:1:8.538 –> 0:1:24.218 Corey Milliman To start copying and pasting out a copilot’s biz chat into other platforms. So it gives us that seamless integration between the entire stack and allows for real time ideation and content creation across teams. 0:1:25.218 –> 0:1:27.738 Corey Milliman So we get that real time collaboration. 0:1:27.978 –> 0:1:34.978 Corey Milliman We have copilot and our AI assistant built into those tasks, so we’re going to go through this live here after we go through a couple of slides. 0:1:35.768 –> 0:1:40.48 Corey Milliman And this is going to also help us with some task management process management. 0:1:40.808 –> 0:1:47.368 Corey Milliman And allows us actually to start reusing some of these components we’re creating with copilot across the stack here. 0:1:47.368 –> 0:1:49.248 Corey Milliman So I’m actually going to stop. 0:1:50.808 –> 0:1:54.888 Corey Milliman And I’m going to stop the slideshow and. 0:1:56.568 –> 0:2:2.408 Corey Milliman Share my screen out differently. It looks like it stopped sharing as well, so just a moment here. 0:2:6.428 –> 0:2:8.788 Corey Milliman I was going to share my entire screen to make it easier. 0:2:10.568 –> 0:2:13.728 Corey Milliman And we’ll flip back and forth here this way. 0:2:13.728 –> 0:2:15.328 Corey Milliman So what? 0:2:15.328 –> 0:2:19.208 Corey Milliman They what Microsoft has done is in teams. 0:2:19.288 –> 0:2:23.528 Corey Milliman We have this new component that’s been released called caused dischar. 0:2:25.648 –> 0:2:28.968 Corey Milliman GE get some notifications here. 0:2:28.968 –> 0:2:30.568 Corey Milliman Work smarter with copilot. 0:2:30.688 –> 0:2:31.968 Corey Milliman We’re going to copilot here. 0:2:33.768 –> 0:2:36.328 Corey Milliman And this used to be called copilot chat. 0:2:36.328 –> 0:2:38.88 Corey Milliman Now this is called discharged. 0:2:38.328 –> 0:2:42.688 Corey Milliman So anytime I am working with copilot now and I’m generating ideas. 0:2:43.808 –> 0:2:49.328 Corey Milliman Let’s generate 10 ideas based on copilot. 0:2:50.888 –> 0:3:6.488 Corey Milliman Wave to training activities so once I start working with copilot and start asking it questions right, it’s going to come back and it’s going to look through my internal documents that I’ve accessed to and it’s giving me some different ways that we can do this and some differe. 0:3:6.528 –> 0:3:14.528 Corey Milliman Ideas so directly from here, instead of copying this over, bring it over into an e-mail. I can actually start clicking edit in pages. 0:3:15.268 –> 0:3:30.708 Corey Milliman And this is going to allow me to create loop components that I can start sharing with other individuals in the organization, turning these into components that I could reuse across other applications and start generating different types of activities here. 0:3:30.908 –> 0:3:40.308 Corey Milliman So if I decided that we are going to maybe have some peer learning groups, I want to develop a class on this and agenda. 0:3:41.168 –> 0:3:42.888 Corey Milliman I’m going to go back into Co filer. 0:3:42.888 –> 0:3:45.88 Corey Milliman Let’s work on an agenda. 0:3:46.498 –> 0:3:47.338 Corey Milliman This workshop. 0:3:50.808 –> 0:3:53.608 Corey Milliman And I’m gonna close this out of pilots over here. 0:3:53.608 –> 0:3:54.768 Corey Milliman We’re gonna let this go. 0:3:54.768 –> 0:3:57.248 Corey Milliman It’s going through some documents that I’ve written before. 0:3:57.688 –> 0:4:13.448 Corey Milliman It’s going through a training workshop agenda for me, and now I’m going to use this and say this is the agenda that we’re going to use for workshop and I’m going to go ahead and turn this into a component that I can start using them across the entire. 0:4:13.548 –> 0:4:15.948 Corey Milliman Their office stack so I can bring this over into loop. 0:4:16.308 –> 0:4:20.628 Corey Milliman I can start sharing this with other individuals as I make changes. 0:4:20.628 –> 0:4:25.748 Corey Milliman Here these different components are going to start changing across the stack. 0:4:25.948 –> 0:4:40.348 Corey Milliman So if I copy this into PowerPoint or copy it into word, or I bring these components into outlook, all of these AI driven insights that I’m generating through biz chat are going to start showing up everywhere I’ve copied up. So as I work with other individuals in the. 0:4:40.348 –> 0:4:41.468 Corey Milliman Organization right and. 0:4:42.208 –> 0:4:43.88 Corey Milliman Amy is on her call today. 0:4:43.168 –> 0:4:44.128 Corey Milliman She’s part of our marketing department. 0:4:44.848 –> 0:5:0.528 Corey Milliman If we’re working on a new agenda, then Amy could also bring some of her insight. Some of her documents, some of her organizational knowledge into the conversation using copilot, and then together we could have this agenda put together pretty quickly and then brought back over into something that. 0:5:0.528 –> 0:5:1.48 Corey Milliman We can start consuming. 0:5:2.608 –> 0:5:13.488 Corey Milliman So this is just one way that pages has been brought in to copilot to help us have a central and easier way to start sharing what we are actually creating. 0:5:14.218 –> 0:5:15.658 Corey Milliman Across copilot there. 0:5:15.818 –> 0:5:23.18 Corey Milliman So I’m actually gonna pause and I’m going to see if there are any questions or thoughts so far before we kind of continue. 0:5:26.818 –> 0:5:29.858 Corey Milliman No questions or thoughts. Alright then. 0:5:31.408 –> 0:5:42.288 Corey Milliman The two parts that I think are that are going to be more valuable today actually are going to be when we get into copilot and Python in Excel and actually creating copilot agents. 0:5:42.288 –> 0:5:50.848 Corey Milliman So I’m going to go back over to our slide deck and I’m just going to maximize these slides because I’m actually going to be bouncing back and forth and it’s a little bit easier here. 0:5:52.648 –> 0:5:56.688 Corey Milliman So Python is now integrated into Excel. 0:5:57.618 –> 0:6:7.858 Corey Milliman So what that means is if you’re working with a lot of data, Excel has always been the tool we’re always going to use for financial ADV analysis or reporting. Any kind of data manipulation. 0:6:8.458 –> 0:6:17.178 Corey Milliman But when we’re actually performing more advanced kind of data analytics, Python has much more powerful libraries for data science. 0:6:17.618 –> 0:6:26.578 Corey Milliman So with that being directly built into Excel, we can start using Python’s data analysis libraries like pandas for data manipulation. 0:6:27.218 –> 0:6:36.178 Corey Milliman Or other tools for visualization without having to create a document, export it from Excel, bring it into Python, and then bring it back into Excel. 0:6:36.938 –> 0:6:53.88 Corey Milliman So when we are looking at how we clean up our data, how we format it, how we visualize trends, we’re relying really heavily on all these charts. But with that integration, we can bypass a lot of the limitations that we have in Excel and start getting into a. 0:6:53.88 –> 0:6:53.938 Corey Milliman More advanced analysis. 0:6:54.728 –> 0:7:0.88 Corey Milliman So here I actually use copilot to create a table for us today with some data. 0:7:0.498 –> 0:7:1.658 Corey Milliman That we can use for this demo. 0:7:3.418 –> 0:7:8.298 Corey Milliman And access Python in copilot there are a couple of different ways you can do this. 0:7:8.618 –> 0:7:25.938 Corey Milliman Going to expand this out here so you can see the easiest way is to get into is to click advanced analysis and this is actually going to start an interaction with copilot that allows us to start asking questions about our data. Copilot is going to ask US quest. 0:7:26.18 –> 0:7:27.858 Corey Milliman About what we’re trying to achieve. 0:7:28.648 –> 0:7:34.328 Corey Milliman And we can actually go back and forth using copilot to write Python, add our own Python code to different cells. 0:7:35.28 –> 0:7:38.788 Corey Milliman And do another other things like that to perform advanced analysis. 0:7:39.108 –> 0:7:42.188 Corey Milliman So I’m gonna go ahead and click start advanced analysis. 0:7:45.28 –> 0:7:48.548 Corey Milliman You can see here it’s actually created a new sheet for me. 0:7:48.828 –> 0:7:57.948 Corey Milliman It’s actually going through and starting to run an analysis. For me, it’s showing the Python code that it’s generated. In order to do this analysis. 0:8:1.308 –> 0:8:4.828 Corey Milliman And you can see here it’s starting to put out different pieces of information. 0:8:4.828 –> 0:8:13.908 Corey Milliman So I have a list of some copilot code that I’ve actually put together that I’m going to go pull up in another document here. 0:8:14.548 –> 0:8:17.108 Corey Milliman So some different things that we can start doing with this. 0:8:17.548 –> 0:8:25.468 Corey Milliman So we can have copilot import different pandas or different items into different libraries to help us with our data analysis. 0:8:26.328 –> 0:8:30.568 Corey Milliman And we can perform these data analysis with real, you know, real language. 0:8:31.378 –> 0:8:37.458 Corey Milliman So maybe I want to filter my data set to show rows where the region is north and sales volume is greater than 100. 0:8:38.298 –> 0:8:48.338 Corey Milliman So it’s actually going to translate that into Python code for me, and it’s going to load that data set. It’s showing me the code that it inserted and go back here and it’s showing me it. 0:8:48.338 –> 0:8:50.58 Corey Milliman Put that in cell a 24. 0:8:52.58 –> 0:8:53.818 Corey Milliman And now it’s scrolling through. 0:8:53.818 –> 0:9:8.298 Corey Milliman So if I go back over to a 24, we can actually see that I actually it shows PY which shows that that’s actually Python And we have these different comments here. The actual Python code right in the cell. 0:9:8.538 –> 0:9:12.578 Corey Milliman So I can start interacting with this if I was a Python developer, I could start making changes here. 0:9:12.858 –> 0:9:16.538 Corey Milliman I could also ask copilot to start doing different things for me. 0:9:17.328 –> 0:9:22.608 Corey Milliman With that information, so I can actually also say create a line pot. 0:9:23.678 –> 0:9:25.158 Corey Milliman Create some visualizations. 0:9:25.918 –> 0:9:29.78 Corey Milliman So let’s create a line plot for sales volume over time. 0:9:29.878 –> 0:9:34.998 Corey Milliman It’s actually going to import the right library. For me. It’s the math plot library. 0:9:34.998 –> 0:9:41.678 Corey Milliman It’s going to bring that over. Start understanding the data and you can see here it used Python. 0:9:41.878 –> 0:9:45.38 Corey Milliman I keep scrolling over here to generate this chart. 0:9:45.38 –> 0:9:52.358 Corey Milliman We’ll have to let it finish now that it’s done, we can see I have a chart here that shows sales volume over time. 0:9:53.208 –> 0:9:58.848 Corey Milliman And we also can use advanced charting if we have really large data sets and start using seaborne. 0:9:59.608 –> 0:10:5.248 Corey Milliman Or other types of libraries to visualize data a little bit differently. 0:10:6.808 –> 0:10:21.168 Corey Milliman So we could also scatterplots or something that is done with seaborne and it’s going to automatically import that library for us and start creating that visualization for us here by revenue versus sales volume colored by region. 0:10:24.798 –> 0:10:33.78 Corey Milliman So you can see that we can get a lot of insights using real language and still have that powerful Python data analysis behind us. 0:10:33.318 –> 0:10:39.718 Corey Milliman Now we’re not limited just to excel formulas or things like that. Once we get into data analysis. 0:10:40.438 –> 0:10:41.158 Corey Milliman You can see here. 0:10:41.158 –> 0:10:53.798 Corey Milliman Now copilot is starting to understand some of my requests. So it’s showing. Suggest more ideas offering different things it can do. And of course it’s looking at the data that we have here. 0:10:54.748 –> 0:11:10.348 Corey Milliman And it’s saying here’s some different ways we can analyze our data. And so we’re going to go ahead. And if we’re following through and listening to these prompts and having this conversation back and forth, it’s easier for us to get to that end result of visualizing or working. 0:11:10.348 –> 0:11:12.788 Corey Milliman With our data, depending on what our end goal is. 0:11:13.588 –> 0:11:20.748 Corey Milliman So to statistical summaries advanced visualization, there are a lot of different ways that we can start using. 0:11:21.528 –> 0:11:26.128 Corey Milliman Python directly in Excel instead of having to again export that data out. 0:11:26.818 –> 0:11:28.938 Corey Milliman Transform it and then bring it back here. 0:11:30.488 –> 0:11:38.88 Corey Milliman So you can see now it’s giving me some other pieces of information here showing me the correlation and coefficient. 0:11:38.88 –> 0:11:39.448 Corey Milliman And this gets pretty deep here. 0:11:39.448 –> 0:11:55.568 Corey Milliman So this might not might not be something that we’re all using everyday, but the advanced analysis is going to start using advanced features and advanced code to give us different types of insights into our data than what we would get just by interacting with excel’s formulas and. 0:11:55.568 –> 0:11:55.768 Corey Milliman Spreadsheets. 0:11:56.698 –> 0:12:9.698 Corey Milliman So now I come back over here and I’ve exited that kind of advanced analysis mode and now I have the ability to start interacting with copilot and Excel in a way that might be something we’re a little bit more used to. 0:12:12.88 –> 0:12:15.888 Corey Milliman So we can also start doing some things like this when. 0:12:17.488 –> 0:12:22.248 Corey Milliman Some different things with this. When it starts coming into different process automations and different types of things. 0:12:24.48 –> 0:12:26.288 Corey Milliman So one way this is really powerful. 0:12:26.818 –> 0:12:33.338 Corey Milliman Across the stack, when we start talking about pages in Excel and all the different ways we use copilot in office. 0:12:34.888 –> 0:12:48.688 Corey Milliman Is, I think, the biggest thing that people can take away from today is this new thing called copilot agents. And I’m going to bring back up our slide deck and we’re going to talk through agents and then we’re actually going to create a no code agent that’s going. 0:12:48.688 –> 0:12:54.528 Corey Milliman To do something for us on today’s call, because I think that’s where we’re going to have the most value out of everything. 0:12:54.528 –> 0:12:57.848 Corey Milliman Everything else is available kind of as reference material. 0:12:58.818 –> 0:13:8.378 Corey Milliman We do have the Python code in the deck, but copilot agents are really important because they allow us to start automating task and taking actions. 0:13:8.378 –> 0:13:20.458 Corey Milliman So we are actually putting the ability to create an agent with the person who is most familiar with the data or the piece of information they’re working with. So there’s no code. 0:13:21.748 –> 0:13:26.68 Corey Milliman There is the ability to start customizing these with copilot studio. 0:13:26.438 –> 0:13:28.918 Corey Milliman In my environment, that ability is not there yet. 0:13:28.918 –> 0:13:44.198 Corey Milliman That is a feature that is rolling out so we can use this as a baseline. And again, this is now called Biz Chat. When we’re in teams, this is no longer copilot chat, so it’s going to help us automate some of these repetitive tasks. When you think about. 0:13:44.198 –> 0:13:47.798 Corey Milliman The way we use chat bots in AI, it can help us improve our efficiencies. 0:13:48.78 –> 0:13:56.198 Corey Milliman It’s giving me a personalized user experience as long me basically to use copilot to write an app to do something without having to. 0:13:56.568 –> 0:13:57.248 Corey Milliman Out of code. 0:13:57.848 –> 0:14:4.408 Corey Milliman So really wait a really easy way to do this is to go back to copilot and we’re doing new chat here. 0:14:5.948 –> 0:14:11.508 Corey Milliman And on the right here you see I have the ability to get copilot agents or create agents. 0:14:12.108 –> 0:14:26.148 Corey Milliman So if you click get copilot agents, they’re already libraries of agents out there that Microsoft has put out. Especially these with the copilot icon that are going to help you use different applications across the stack. 0:14:26.898 –> 0:14:29.338 Corey Milliman And these copilot icons mean these are actually. 0:14:30.908 –> 0:14:36.28 Corey Milliman Enhanced to work with Microsoft Copilot, but today we’re actually going to create one. 0:14:36.28 –> 0:14:40.748 Corey Milliman So I’m gonna go back to copilot, and there are a couple of different things that we can do here. 0:14:41.468 –> 0:14:49.68 Corey Milliman So if we create an agent, we actually are going to use AI to describe what we want the agent to do. 0:14:50.788 –> 0:14:57.828 Corey Milliman And then we are actually going to be able to configure data sources that give it those sources of instructions, those sources of knowledge. 0:14:58.838 –> 0:15:3.598 Corey Milliman That could be a SharePoint site, a Word document, a PowerPoint, a website. 0:15:3.798 –> 0:15:13.598 Corey Milliman You know all different things like that, but we can only have up to 20 knowledge sources when we’re doing it this way. And we can also talk about whether or not we want to use web-based content. 0:15:13.598 –> 0:15:21.958 Corey Milliman So are we also using the Internet along with this or are we only grounding this chatbot with the knowledge that we specify here? 0:15:22.798 –> 0:15:28.438 Corey Milliman So one way I’ve done this and then we’re going to go back and build that is I’ve built out what’s a contract review. 0:15:28.958 –> 0:15:29.798 Corey Milliman Assistant here. 0:15:31.348 –> 0:15:34.788 Corey Milliman And this is going to help to redline contracts based on standard knowledge. 0:15:36.468 –> 0:15:39.908 Corey Milliman I actually show you how I built this on the back end. 0:15:42.638 –> 0:15:53.318 Corey Milliman So I have a document that I created that has our some standard contract terms that we wanna see in all of our. In all of our documents and this is not something that is. 0:15:55.308 –> 0:15:58.148 Corey Milliman Proprietary to concurrency or anything like that. 0:15:58.508 –> 0:16:13.308 Corey Milliman These are just some terms that I came up with for today’s demo, so I actually have those loaded on a SharePoint site right now and these are just a list of standard contract terms that we might want to see. 0:16:13.748 –> 0:16:18.588 Corey Milliman I have my standard payment terms, termination clause, confidentiality agreements. 0:16:18.868 –> 0:16:19.588 Corey Milliman I just have a list. 0:16:20.388 –> 0:16:26.268 Corey Milliman I have a hiding of what that term is and then I say it should be net 30 and go through that. 0:16:26.808 –> 0:16:28.168 Corey Milliman And that just has a list. 0:16:28.168 –> 0:16:30.848 Corey Milliman Very simple list to start out with. 0:16:31.648 –> 0:16:34.488 Corey Milliman So then I went ahead and I created my agent. 0:16:34.488 –> 0:16:37.608 Corey Milliman We’re gonna walk through that, but I’m first gonna show you how I use it. 0:16:38.848 –> 0:16:44.8 Corey Milliman So what I can actually do with agents is I can start calling them from other applications. 0:16:44.448 –> 0:16:52.168 Corey Milliman All is like an assistant, so I’m going to bring up a a very simple contract here. And again, this was created for today’s demo. 0:16:52.948 –> 0:16:57.308 Corey Milliman Microsoft Jumpstart project contract I’m going to open up copilot. 0:17:0.78 –> 0:17:12.558 Corey Milliman You can see here I have copilot that we’re familiar with in Word, and I can actually choose different copilots in the chatbot that I put together. My contract review assistant now is over here because this is my job. 0:17:12.558 –> 0:17:15.558 Corey Milliman I’m reviewing contracts. I created this to make my life easier. 0:17:17.108 –> 0:17:19.908 Corey Milliman So now I just want to go through and I want to say. 0:17:21.668 –> 0:17:23.28 Corey Milliman Reference my repository. 0:17:23.28 –> 0:17:26.388 Corey Milliman Reference this repository to ensure this contract is accurate. 0:17:26.828 –> 0:17:33.468 Corey Milliman The repository represents the information I grounded copilot with saying these are my standard terms. These are my standard services. 0:17:34.388 –> 0:17:35.828 Corey Milliman These are my standards. 0:17:36.268 –> 0:17:44.468 Corey Milliman And now it’s going to go through and it’s going to give me a list of all the changes I need to make to this document. So as it goes through, it’s going to expand here in a moment. 0:17:46.788 –> 0:18:2.668 Corey Milliman So if there’s anything that we are doing in our jobs that we are looking for, say, in a Word document where we want standard language or we’re writing a proposal or we’re writing a document, these standards can be stored in a centralized repository that we’re all working on. 0:18:3.68 –> 0:18:12.908 Corey Milliman And then as that repository evolves, now everybody that has this chatbot has a central legal source of information. If we change our payment terms or something like that. 0:18:13.708 –> 0:18:17.668 Corey Milliman So you can see it’s highlighting in this long contract. Everything that needs to change. 0:18:18.278 –> 0:18:32.478 Corey Milliman Payment terms termination clause, confidentiality, intellectual property, and then I have these notations that are bringing me back to that main document which shows where it learned this information from. 0:18:33.158 –> 0:18:41.838 Corey Milliman So I always have that source of truth because I using copilot and responsible for this output and I have to make sure that everything I’m putting together is correct. 0:18:42.518 –> 0:18:45.38 Corey Milliman Now I’m going to pause because I see a chat bubble. 0:18:45.788 –> 0:18:46.628 Corey Milliman And I want to make sure. 0:18:49.438 –> 0:18:55.678 Corey Milliman So when using web-based data sources can use a web portal that requires a login like JIRA. 0:18:56.358 –> 0:19:14.188 Corey Milliman No, not with. So short answer is no, we cannot use something that requires a login when we’re creating a copilot. From a user perspective without any coding, we’re creating this agent directly within copilot. In order to do something like that, we would need copilot studio and we would. 0:19:14.188 –> 0:19:19.118 Corey Milliman Have to create that connector, create that login so that we had access to a different source of inform. 0:19:21.38 –> 0:19:23.38 Corey Milliman So does that answer your question, Daniel? 0:19:25.568 –> 0:19:27.928 Corey Milliman Awesome. All right. So that’s helpful. 0:19:27.928 –> 0:19:29.528 Corey Milliman So yes, that’s a good question though. 0:19:31.268 –> 0:19:36.868 Corey Milliman These are copilot agents that are being pushed down at the user level, and there are some differences. 0:19:38.468 –> 0:19:53.668 Corey Milliman Because there are instances right where we want that chatbot to have access to a proprietary system or different things like that. In copilot Studio, we can set up all these connectors and set up all these different actions that are bring all kinds of information back over to the. 0:19:53.668 –> 0:19:58.988 Corey Milliman Chatbot experience but this chatbot and what it how it relates to copilot? 0:19:59.568 –> 0:20:5.128 Corey Milliman Is also different in the fact that messaging is something that’s different between the two. 0:20:5.128 –> 0:20:16.688 Corey Milliman So when you create a chatbot and copilot studio, you are allowed so many messages per month or you have different quotas and you’re charged, maybe based on the number of message interactions that go through. 0:20:18.228 –> 0:20:24.668 Corey Milliman Our copilot agents that are being published with the copilot license this way do not have any messaging limits. 0:20:24.668 –> 0:20:32.868 Corey Milliman So if I put out something in my organization that is really, really helpful to everybody and I share this chatbot out and everybody starts using it. 0:20:33.518 –> 0:20:45.838 Corey Milliman I don’t have any additional costs other than my copilot license here, so it’s a little bit different than copilot studio where we’re actually metering the number of messages and the number of actions that are taking place. 0:20:46.198 –> 0:20:58.918 Corey Milliman So this truly does allow those individual users that know the data, know their processes, take a no code approach to creating something that’ll help automate part of their day, or take care of part of their workflow. 0:21:0.148 –> 0:21:3.108 Corey Milliman And yes, this is also available for SharePoint. 0:21:4.218 –> 0:21:8.258 Corey Milliman If you go out to a SharePoint site, you will actually see. 0:21:9.828 –> 0:21:15.788 Corey Milliman Let me I’m going to bring up in another window here because I don’t want to bring up any other customer information. 0:21:15.788 –> 0:21:16.468 Corey Milliman So just a moment. 0:21:18.188 –> 0:21:29.948 Corey Milliman There is a copilot icon. You will see next to a file in SharePoint and you can actually click right there and start building your copilot agent from that data source. 0:21:33.138 –> 0:21:39.978 Corey Milliman Just waiting for SharePoint to load in another window and so while that’s loading, I’ll make sure we get back to that here. 0:21:48.318 –> 0:21:48.958 Corey Milliman There we go. 0:21:49.278 –> 0:21:53.878 Corey Milliman Sorry, my entire computer just froze for a moment because I have so many things open for this. 0:21:53.878 –> 0:21:57.238 Corey Milliman So alright, so how do we do this? 0:21:57.238 –> 0:22:10.198 Corey Milliman Let’s create. There’s also another chat bot here that I did create and I’m showing you this other chat bot because I wanna also show something that’s called token limits when we’re working with our creating our bots. 0:22:11.118 –> 0:22:12.118 Corey Milliman I called one. 0:22:12.118 –> 0:22:14.358 Corey Milliman I created one called copilot companion. 0:22:15.148 –> 0:22:16.28 Corey Milliman And what I did here. 0:22:17.828 –> 0:22:19.668 Corey Milliman Was I gave it a set of instructions. 0:22:20.288 –> 0:22:27.768 Corey Milliman Now we’re gonna bring these instructions up in Word, but I’m allowed to write instructions that are up to 8000 characters in length. 0:22:28.848 –> 0:22:45.408 Corey Milliman And again, I have a knowledge source here, so I’m going to bring over what I did so we can kind of talk through the logic that copilot agents use. And as you can see, I have a ton open on my computer and I’m trying not to. 0:22:46.508 –> 0:22:47.388 Corey Milliman Just go this way. 0:22:49.148 –> 0:22:49.468 Corey Milliman All right. 0:22:51.108 –> 0:22:51.788 Corey Milliman So this is. 0:22:52.798 –> 0:23:6.878 Corey Milliman How I trained this agent to do what it’s going to do and what this agent does is I have a set of prompts that collected over the last year, about two hundred 250 high value prompts that are really, really helpful across the different office applications. 0:23:8.428 –> 0:23:22.548 Corey Milliman So I told copilot that it’s going to be called copilot companion, and what it’s actually going to do. It’s a conversational assistant design to help people with Outlook, Word, Excel, PowerPoint, and teams. I told them what tone do you use. 0:23:23.298 –> 0:23:31.458 Corey Milliman And I also said guide users on the actions they wanna perform different comp categories and the Microsoft product they’re using. 0:23:32.828 –> 0:23:34.348 Corey Milliman And then I gave it an example. 0:23:34.468 –> 0:23:41.988 Corey Milliman So say for category called catch up and I use that category because across the stack, whenever using copilot you click. 0:23:43.698 –> 0:23:45.858 Corey Milliman There’s a prompt button over here. 0:23:46.578 –> 0:23:55.178 Corey Milliman View prompts. Generally, Copilot Lab is going to be under certain types of categories where we’re creating content. 0:23:55.178 –> 0:23:58.978 Corey Milliman Asking, summarizing, you know, different things like that. 0:23:58.978 –> 0:24:3.258 Corey Milliman So I gave an example of how I wanted to respond. 0:24:4.598 –> 0:24:13.438 Corey Milliman I told it exactly what I wanted it to do when the user specified their task, I said provide a relevant prompt along with the description explaining how it used, how it’s used. 0:24:14.978 –> 0:24:17.58 Corey Milliman And I said, here’s how to approach each category. 0:24:17.138 –> 0:24:23.538 Corey Milliman So I wrote this document that says if somebody says they want to learn something, provide them prompts to help them learn information. 0:24:25.568 –> 0:24:29.848 Corey Milliman And then I actually went through the user interaction and scenarios and told her how to respond. 0:24:31.418 –> 0:24:49.898 Corey Milliman Gave it a response flow and told it exactly what I wanted it to do. So by writing this up and then putting this in copilot, I’m actually then able to switch over to our copilot companion. And now I have these tiles that I also created without any code. 0:24:50.818 –> 0:24:53.738 Corey Milliman And I told it what these tiles would be. 0:24:54.298 –> 0:24:56.778 Corey Milliman I would like to something near to learn. 0:24:57.318 –> 0:25:3.958 Corey Milliman Something new with copilot, just different ways of prompting users and giving users information. 0:25:3.998 –> 0:25:7.758 Corey Milliman And again, this was no code, so I need some help digging deeper. 0:25:7.758 –> 0:25:13.998 Corey Milliman Can you suggest ask from us now? It’s going through saying now it’s personalized to me. 0:25:13.998 –> 0:25:14.758 Corey Milliman It knows my name. 0:25:14.758 –> 0:25:16.598 Corey Milliman What kind of insights are you looking for? 0:25:16.598 –> 0:25:17.958 Corey Milliman Are you working excel? 0:25:17.958 –> 0:25:20.838 Corey Milliman You know, what are you doing? I’m working. 0:25:22.338 –> 0:25:27.498 Corey Milliman In Microsoft Word and I want to learn more about. 0:25:30.698 –> 0:25:35.898 Corey Milliman A-150 page document without reading it all. 0:25:39.588 –> 0:25:48.468 Corey Milliman So now one limitation with these is this also going to go through and it’s looking at you saw a Flash 150 page document. 0:25:48.708 –> 0:25:58.908 Corey Milliman So initially it was going through and looking through everything I had access to in OneDrive and e-mail and chats looking for 100 and page 50 page document. 0:25:59.388 –> 0:26:6.948 Corey Milliman But now it knows I’m not referencing an explicit document with that title and understands that these are the different things I want to do. 0:26:7.738 –> 0:26:12.378 Corey Milliman So here are some prompts that can help you dig deeper, and these are ways that I can start using this here. 0:26:12.378 –> 0:26:15.578 Corey Milliman So it’s telling me summarize the key points, create an outline. 0:26:16.608 –> 0:26:30.968 Corey Milliman So I created this so that different people across the organization would have a way of getting different ways different ideas about how to use copilot and then like I said, we also had the one where we looked at today about contract review. 0:26:31.488 –> 0:26:40.848 Corey Milliman So if we have any document that we are using as a sort of source of truth, we can create an agent to do something with that document. 0:26:40.848 –> 0:26:43.288 Corey Milliman And now we’re going to recreate our legal chat bot today. 0:26:44.778 –> 0:26:44.938 Corey Milliman So. 0:26:46.538 –> 0:26:48.138 Corey Milliman I so it’s asking me. 0:26:49.348 –> 0:26:52.228 Corey Milliman Exactly. To describe what I wanted to do. 0:26:53.818 –> 0:27:0.778 Corey Milliman And in real language I want to create a chatbot that will help me identify. 0:27:2.418 –> 0:27:6.18 Corey Milliman Standard contract language deviations. 0:27:6.338 –> 0:27:9.58 Corey Milliman The chatbot should be friendly. 0:27:10.858 –> 0:27:13.658 Corey Milliman And help users identify. 0:27:15.258 –> 0:27:16.858 Corey Milliman Contract issues. 0:27:20.168 –> 0:27:33.768 Corey Milliman And you can get much longer with these. So I’m trying to keep this short for our demo today, but with copilot, the more direction I give it, the more context I give it. The better examples I give it, the more effective it’s going to be, so saying great. 0:27:33.768 –> 0:27:37.648 Corey Milliman OK, your copilot’s going to helps identify standard contract language. 0:27:37.648 –> 0:27:41.528 Corey Milliman Do you have any specific instructions on how copilot should exist? 0:27:41.528 –> 0:27:45.288 Corey Milliman Like a specific tone and just for damaged no more instructions. 0:27:50.528 –> 0:27:54.888 Corey Milliman And while it’s going ahead and doing that, I’m going to give it a name. 0:27:56.458 –> 0:28:3.338 Corey Milliman And I’ll say contract 2, this is a description that we’re going to write up so that users can understand how to use this. 0:28:3.858 –> 0:28:10.458 Corey Milliman The instructions are actually the instructions on what the copilot agent is going to do, written in natural language. 0:28:10.458 –> 0:28:14.298 Corey Milliman And then we’re going to browse over to our knowledge source. 0:28:15.938 –> 0:28:20.738 Corey Milliman And I am going to. I can’t bring this into going to bring this into another window. 0:28:23.318 –> 0:28:26.998 Corey Milliman And I’m going to select my knowledge source and then I’ll bring it back over. 0:28:29.838 –> 0:28:30.518 Corey Milliman Here we go. 0:28:36.28 –> 0:28:37.708 Corey Milliman We’re gonna have to do this live on the screen. 0:28:37.868 –> 0:28:38.428 Corey Milliman Just a moment. 0:28:42.348 –> 0:28:45.28 Corey Milliman So here I already have this standard contract terms. 0:28:45.28 –> 0:28:46.548 Corey Milliman I uploaded this to a team site. 0:28:46.548 –> 0:28:49.988 Corey Milliman It can use any SharePoint document. I could use a library. 0:28:49.988 –> 0:28:54.748 Corey Milliman I could use many folders and now I’ve grounded that with knowledge. 0:28:55.668 –> 0:29:7.828 Corey Milliman So here I mentioned that actions and other capabilities will be coming soon with Copilot Studio, where I could take this this workflow, this, this copilot agent that I’ve already built. 0:29:8.618 –> 0:29:10.658 Corey Milliman Open it up in copilot studio and then extend it. 0:29:10.658 –> 0:29:16.418 Corey Milliman So maybe it could go pull information from JIRA or start pulling information from a personalized knowledge source. 0:29:17.268 –> 0:29:20.628 Corey Milliman But this is my, you know most basic starter version. 0:29:22.178 –> 0:29:29.698 Corey Milliman So now it actually went through and it’s it came up with some suggestions on starter prompts and these are these tiles that we see on the right hand side. 0:29:30.18 –> 0:29:35.698 Corey Milliman So saying identify contract deviations review language highlight issues, compare clauses. 0:29:35.818 –> 0:29:43.538 Corey Milliman So I’m just going to go ahead and accept these defaults and then I have the ability from here to either start testing this. 0:29:45.258 –> 0:29:48.458 Corey Milliman Or I can go out and I’m going to actually create this. 0:29:49.318 –> 0:29:54.678 Corey Milliman And now it’s creating that copilot and publishing it to my organization so that I can start using this. 0:29:54.718 –> 0:29:56.518 Corey Milliman And I called this one contract. 0:29:58.98 –> 0:29:58.218 Corey Milliman 2. 0:30:1.18 –> 0:30:5.778 Corey Milliman It does tell you do not close the window because if you do do that, it will not work. 0:30:6.378 –> 0:30:14.658 Corey Milliman Will bomb during the creation of that you do have the ability immediately to share this out to other individuals in your organization just by sharing out this link. 0:30:16.458 –> 0:30:18.858 Corey Milliman But I’m actually right now I have it set. 0:30:18.858 –> 0:30:22.258 Corey Milliman It’s only for me ’cause. I don’t want to make this public yet because I haven’t. 0:30:22.258 –> 0:30:24.58 Corey Milliman Actually, I haven’t even tried it yet. 0:30:24.58 –> 0:30:33.498 Corey Milliman I haven’t done anything with it, so if I click go to agent you can see contract to shows up here in copilot under my list of different agents here. 0:30:34.228 –> 0:30:38.188 Corey Milliman So these are all agents that I put together myself and they’re all under my list. 0:30:38.828 –> 0:30:41.828 Corey Milliman The reason that these are below this line, these are private. 0:30:41.828 –> 0:30:46.828 Corey Milliman These are my agents only, whereas these are agents that publish to the entire organization. 0:30:48.378 –> 0:30:54.738 Corey Milliman So to make it easier, I’m actually going to go into word because I already have a contract in one of these windows. 0:30:54.738 –> 0:31:4.858 Corey Milliman Here’s another basic contract and I just put out terms that I knew would explicitly wouldn’t work and would be flagged by this tool. 0:31:5.698 –> 0:31:6.138 Corey Milliman Go back. 0:31:6.138 –> 0:31:7.338 Corey Milliman Open up copilot. 0:31:8.628 –> 0:31:13.868 Corey Milliman I’m going to go back to contract 2, the one that we just created. 0:31:15.418 –> 0:31:19.178 Corey Milliman And I’m actually going to say highlight contract issues. 0:31:20.938 –> 0:31:23.18 Corey Milliman So now it’s going to go back to that Word document. 0:31:24.738 –> 0:31:27.858 Corey Milliman And it’s actually now using that as the source. 0:31:29.418 –> 0:31:30.58 Corey Milliman We’ll wait for that to finish. 0:31:34.418 –> 0:31:39.458 Corey Milliman You can see here, it’s actually said it’s actually identified those same terms. 0:31:39.938 –> 0:31:46.98 Corey Milliman So now if somebody were to go back and change that source document, that source of truth and update our payment terms. 0:31:48.58 –> 0:31:49.498 Corey Milliman Like here I have this open. 0:31:51.178 –> 0:31:57.498 Corey Milliman Now I’m going to change this to 90 days. I’m going to say either party can terminate with 180 days notice. 0:31:59.138 –> 0:31:59.698 Corey Milliman And hopefully. 0:32:1.418 –> 0:32:2.418 Corey Milliman This saved just now. 0:32:3.988 –> 0:32:8.628 Corey Milliman And it does take a moment for copilot to understand that that document has changed. 0:32:8.628 –> 0:32:10.268 Corey Milliman So we’re gonna see how fast it happens today. 0:32:12.698 –> 0:32:14.898 Corey Milliman And switch back over to contract 2. 0:32:16.578 –> 0:32:21.218 Corey Milliman And see if it already understands the new information that I put out there. 0:32:21.778 –> 0:32:29.938 Corey Milliman Usually I see this take 5 to 10 minutes, but we’re gonna try it live since we have a group of people on the phone and just see what happens. 0:32:31.578 –> 0:32:38.98 Corey Milliman So while that is doing this, are there any questions or thoughts on how we can use copilot agents across the organization? 0:32:38.578 –> 0:32:39.378 Corey Milliman Other use cases. 0:32:45.548 –> 0:32:48.228 Corey Milliman And whatsoever easy group today. 0:32:51.468 –> 0:32:52.788 Corey Milliman So that’s still looking through. 0:32:52.788 –> 0:32:58.388 Corey Milliman I have a feeling this is not going to finish until I close the document and reload just because I changed I copilot agent. 0:32:59.938 –> 0:33:3.58 Corey Milliman So I’m going to go back over. I do see. 0:33:4.698 –> 0:33:8.778 Corey Milliman Let me go back over to our PowerPoint and just make sure I didn’t ignore anything. 0:33:9.138 –> 0:33:11.818 Corey Milliman But there are some other ways that you can use those copilot agents. 0:33:11.818 –> 0:33:14.818 Corey Milliman So is this running locally or in the cloud? 0:33:14.818 –> 0:33:22.338 Corey Milliman This is actually running in the cloud that copilot agent is associated with my own Microsoft 365 account and with our tenant. 0:33:22.858 –> 0:33:25.298 Corey Milliman So right now it’s only available to myself. 0:33:26.338 –> 0:33:29.418 Corey Milliman And how else can the agents help project managers? 0:33:30.618 –> 0:33:35.258 Corey Milliman There are a lot of different ways, I think umm, from a project management standpoint, agents could be helpful. 0:33:36.818 –> 0:33:43.858 Corey Milliman Basic agents can help us identify when people are available for meetings, what calendars look like. 0:33:43.858 –> 0:33:59.738 Corey Milliman I can also write an agent to if I know that if I have the name of a project that’s out there, say project Phoenix and I have had 30 meetings over Project Phoenix over the last 90 days, I could actually go into copilot and into an agent, say. 0:33:59.738 –> 0:34:1.818 Corey Milliman Write me a status report based on. 0:34:2.468 –> 0:34:4.908 Corey Milliman This type of information was in meetings. 0:34:5.148 –> 0:34:8.28 Corey Milliman Was it e-mail notes? 0:34:8.148 –> 0:34:22.508 Corey Milliman You know, I can have it go back through and comb all those multiple data sources and I could set up a separate agent for each one of my projects so that I was actually maybe creating a status report for Project Phoenix and then another agent to create a. 0:34:22.508 –> 0:34:26.348 Corey Milliman Status report for a different project. I’m on and different things like that. 0:34:26.868 –> 0:34:31.108 Corey Milliman The need for governance of public agent publishing seems necessary. 0:34:31.108 –> 0:34:31.788 Corey Milliman Do you have thoughts? 0:34:33.808 –> 0:34:34.248 Corey Milliman So. 0:34:35.818 –> 0:34:36.738 Corey Milliman You’re absolutely correct. 0:34:36.738 –> 0:34:40.218 Corey Milliman Now one thing about these agents though, is they are. 0:34:42.58 –> 0:34:43.898 Corey Milliman Only we have a couple of questions going on here. 0:34:44.138 –> 0:34:49.778 Corey Milliman Visual creator is actually an agent that I downloaded from Microsoft that allows me to create different visuals. 0:34:50.538 –> 0:34:58.218 Corey Milliman This is a way of doing it without having to open up copilot on the web and create different graphics or image generation there. 0:35:0.58 –> 0:35:2.138 Corey Milliman But from a governance perspective, that is a good question. 0:35:2.808 –> 0:35:3.528 Corey Milliman Because. 0:35:5.98 –> 0:35:22.218 Corey Milliman We do wanna make sure that we’re not exposing an unnecessary risk. However, these copilot agents me as a user of copilot. I can only run this agent against all the information I already have access to. So that’s gonna be my emails, my OneDrive, my conversations, and those other. 0:35:22.218 –> 0:35:29.978 Corey Milliman Pieces of information that I already have access to in my tenant. If I share this copilot agent with somebody else in my organization. 0:35:31.58 –> 0:35:34.978 Corey Milliman It’s only going to work for them if they have access to that source knowledge. 0:35:35.508 –> 0:35:43.268 Corey Milliman So what we’ve done by creating an agent is we do have two points of consideration. We have to consider the governance around the grounding. 0:35:43.268 –> 0:35:52.428 Corey Milliman Documents are knowledge base and we also have to consider the governance and security around the actual publication of this actual chatbot. 0:35:52.588 –> 0:36:3.228 Corey Milliman Want to make sure the correct people do have access to this, and I think that’s why it’s really important that whenever rolling out copilot, you know concurrency. We do make sure that we are going through all of those governance activities. 0:36:4.18 –> 0:36:6.18 Corey Milliman Helping to identify who has access to what. 0:36:6.18 –> 0:36:9.578 Corey Milliman Making sure that copilot is not going to inappropriately. 0:36:10.388 –> 0:36:14.268 Corey Milliman Expose information to the wrong people, so that is. 0:36:14.588 –> 0:36:19.228 Corey Milliman But when we talk about public agents, these are not public. 0:36:19.228 –> 0:36:26.308 Corey Milliman These are only inside of our tenant. When we get into public agents, that is a whole nother conversation. 0:36:26.308 –> 0:36:30.908 Corey Milliman I could do a whole another webinar on with copilot studio and a public facing agent. 0:36:31.188 –> 0:36:32.268 Corey Milliman How to secure that? 0:36:32.268 –> 0:36:33.628 Corey Milliman How to secure the data behind it? 0:36:36.168 –> 0:36:38.8 Corey Milliman I’m just catching up here. 0:36:38.8 –> 0:36:43.928 Corey Milliman I’m sorry, did that kind of answer your question? When we talk about governance and and? 0:36:47.778 –> 0:36:48.18 Corey Milliman OK. 0:36:48.298 –> 0:36:49.818 Corey Milliman Very good scrolling back up. 0:36:51.378 –> 0:36:52.378 Corey Milliman Calendar availability. 0:36:52.378 –> 0:36:59.298 Corey Milliman Yeah, it does help with the calendar availability when you get into once we have these basic agents. 0:37:1.18 –> 0:37:17.178 Corey Milliman Our IT team can actually expand these and copilot studio and add additional actions and work flows and things like that, but this gives us a really good basis, so a lot of help in generating content, pulling information and as these continue to evolve, these can start taking AUT. 0:37:17.378 –> 0:37:17.978 Corey Milliman Actions. 0:37:18.738 –> 0:37:33.458 Corey Milliman So I can actually start writing up agents that are going to maybe monitor my inbox for me and do different things like that to surface information that’s most relevant to my job, to go beyond even, you know, using copilot and outlook to, say, show me what’s important or. 0:37:33.458 –> 0:37:38.938 Corey Milliman Show me what I missed on this project, so I think the thing to remember with these agents is they cannot. 0:37:39.18 –> 0:37:42.978 Corey Milliman We cannot only use them just to do something when I call on them. 0:37:43.418 –> 0:37:48.378 Corey Milliman We can also create agents that are kind of running in the background that are helping us in our day-to-day tasks. 0:37:52.168 –> 0:37:54.888 Corey Milliman So any other questions on anything we’ve covered today? 0:38:0.908 –> 0:38:6.828 Corey Milliman Alright, switch back over to the deck and make sure that I didn’t. 0:38:8.338 –> 0:38:8.978 Corey Milliman Leave anything out here. 0:38:12.368 –> 0:38:16.528 Corey Milliman See somebody else did ask about what do these look like? 0:38:16.648 –> 0:38:19.248 Corey Milliman Can you show how the agents are seeing enterprise wide? 0:38:20.818 –> 0:38:23.938 Corey Milliman So when I go to get an agent. 0:38:27.448 –> 0:38:28.888 Corey Milliman See here I can hear my laptop. 0:38:28.888 –> 0:38:47.8 Corey Milliman I think it’s probably about 100° too hot right now, but you can see here that these apps all these different agents that are available across the organization. I have one that says manage my apps, but those copilots that I’ve shared out individually, they actually aren’t going. 0:38:47.8 –> 0:38:47.408 Corey Milliman To show up here. 0:38:47.408 –> 0:38:49.448 Corey Milliman I haven’t published them to a store. 0:38:49.448 –> 0:38:52.928 Corey Milliman I haven’t done anything like that. When I create my copilot. 0:38:53.778 –> 0:38:57.338 Corey Milliman Bots. I’m sharing those with specific individuals or specific teams. 0:38:58.168 –> 0:39:3.8 Corey Milliman Or putting those in places where only certain individuals have access to. 0:39:3.8 –> 0:39:7.928 Corey Milliman So maybe I drop it on SharePoint site where I’m controlling who has access to that SharePoint site or something there. 0:39:11.18 –> 0:39:12.618 Corey Milliman So I hope that answered that. 0:39:12.618 –> 0:39:19.778 Corey Milliman So I think as a follow up, the only other thing you know we do, we have talked a little bit about. 0:39:21.538 –> 0:39:38.978 Corey Milliman Our you know, governance approaches and some different considerations I think are really important when we’re talking about agents through our partnership with Microsoft, we do have the ability to provide a Microsoft funded copilot value discovery that can really help us make sure that we’re getting into the we. 0:39:39.98 –> 0:39:40.778 Corey Milliman On how we can get copilot out there. 0:39:40.778 –> 0:39:43.818 Corey Milliman How we can implement it and what that will look like for you? 0:39:44.558 –> 0:39:57.198 Corey Milliman And then we also can offer a a complementary, you know, initial assessment and talk about you know what it would take to get copilot in your environment in a secure, secure way without exposing all that risk. 0:39:58.948 –> 0:40:6.468 Corey Milliman So if there are no other questions today or no additions from anybody else, I thank you for your time. And if you have any other questions. 0:40:8.18 –> 0:40:9.658 Corey Milliman Let us know. Thank you so much.