 Hey everybody, this is Christian Buckley doing another MVP Buzz Chat interview and I'm talking today with Matt. Hello. Christian, how are you doing? I'm doing well. How are things there? Very well. Thank you. I love you. You've got the... It's like what background is that? It's the first Zoom background that I found. All right. There we go. Yeah. Yes. So a lot of you out there might be recognizing that background. I don't know what city that is. I recognize some of Microsoft's generic backgrounds from like Seattle area, the Puget Sound, but that one I don't recognize. Yeah. Yeah. Yeah. Little secret for you. I actually work in my master closet. I call it the cloth office. And so that's why I always use virtual backgrounds because it's not a great backdrop for... Very nice. Very nice. Well, yeah. Well, hey, that's... There's pros and cons to using the real versus virtual, but... Well, for folks that don't know you, Matt, who are you, where are you, and what do you do? Yeah. So Matt Jemison, I am based out of Indianapolis, Indiana. And as far as what I do right now, I am the head of intelligence solutions and Microsoft 365 apps at Decada Pharmaceuticals. And what that means in a nutshell is I'm working a lot with Power Virtual Agents, doing a lot with generative AI and Azure Open AI. And then my team is also responsible for custom Microsoft 365 development. So very cool. Well, you know, so I've interviewed a couple of people and that's for the other half of my podcast that are in the healthcare industry and talking about a lot of what they're doing with AI. They're just a natural fit for that kind of... For the complexity of the data that pharmaceutical company owns and manages just makes sense that you to automate that, that you check in the AI stuff. Absolutely. Yeah. And we're going through the process of making sure we're doing it the right way, the right way, aligning with our core values and being really strategic about that. But it is a great time to be a part of technology right now. A lot of exciting new things to play with, really every day or every week. It seems like there's something new. It's funny looking at, and when you write a lot, when you talk a lot, when you interview a lot and then you really pay attention to like all of the most recent announcements. And it's actually been kind of slow the last couple of months. And if you know the Microsoft ecosystem, of course, Microsoft people are on vacation the entire month of August. There's generally like shuffle, people are promoted, there's reorgs that kind of happen. And so it's a slower news cycle. There is tons that's going on, except within the Microsoft ecosystem in the last month. Yeah. Yeah. And I know there's a lot of excitement about the upcoming Power Platform conference. And then before you know it will be at Ignite. And so there's definitely a lot of big hitters for the rest of the year that I'm looking forward to as well. Well, what was your journey to becoming an MVP? So what's your origin story? Yeah, no, it's an interesting one. So I, in 2008, went to a company that was looking for a .NET developer that was interested in learning SharePoint Moss 2007. And so I got fully immersed into that experience, building WSPs, feature receivers, event receivers, packaging it all up and took those .NET skills and worked out really well. And so worked there for about two and a half years. And then I moved on to consulting for the first time in my life. I would remain at that company for 12 years. And I was doing the SharePoint Saturdays really in the Midwest. So I spoke at several in Ohio, Indiana and Illinois having a blast. But my career kept taking off in the consulting world. And so I kept year by year getting a little bit further away from the keyboard and got to a point where I was the managing director of a Microsoft consultancy at this firm. And so I was no longer getting to really play with the technology like I had in the past. So I made a tough decision a year ago that I wanted to make a change. I wanted to get back to the keyboard. I wasn't ready to hang it up. And so I actually joined to Cato Pharmaceuticals, like I said, in that position to help them lead their Power Virtual Agents work and really haven't looked back. And I would say it's really reengaging in the community, getting back to being committed to blogging, to user forums, to speaking, just helping others. Right. I love being able to help someone else that's looking at a problem. They're not sure how to solve it. I think that oddly gets me excited when I have a problem. Like I love that challenge. And so it's really been an opportunity to get closer to the technology and then get back to sort of what I was doing. Like I said, back in the early 2010s with SharePoint Saturdays and whatnot and was fortunate enough to get selected in July and just having a blast with it. That's great. And as we were talking before we started recording, so you also had the benefit of a lot of MVPs within your org. And so with that, I'm assuming that Takeda is pretty supportive of MVPs, recognizes the value of having MVPs on staff and let you play in the community. Yeah, absolutely. There are a lot of great MVPs there. Dave Feldman, who runs our group, the Intelligent Workplace, extremely supportive, Darcy Reed, and just Chris Kent, right? Who's also on my team. They were super supportive and it was great to sort of have people to chat with about what it looks like and what it takes. And so I think that having those mentors and coaches has been great. So what are kind of your passion topics now? What are you speaking on, presenting about, writing about? And also, what are your main contributions? Are you more a writer, presenter? Yeah, I mean, at the end of the day, I do love to code. I've got that background, right? But obviously working more into Power Platform, it's getting into more no-code, low-code. Heavily invested in Power Virtual Agents, heavily invested in Power Automate, right? By extension with that. And then really, really doing a lot with Azure Open AI. So I've done some of that through those mechanisms, done some of that through Azure Functions. I've done some work with Prompt Flow. That's some of the additional topics I really want to get into the blog. And I think Azure Open AI, there's just a lot of opportunity there looking at what's coming, what we have already. And so just trying to understand how to create that generative AI journey for everyone. So I would say that's where I'm really most passionate. I'm actually going to be in Boston next month at CLAB, New England Day is talking about prompt engineering. So it's going to be an introduction to that, just to understand it's really the foundation right of utilizing generative AI. So that's the next convo I've had. And I've done a lot with Azure Open AI PVA type topics as well. You know, we hear it. So obviously you heard a lot, you hear a lot about, there's a lot of utilities out there, courseware, the lot of content out there around building virtual agents and of course everything else in Power Platform. You also, you're more and more content people talking about co-pilot. But what about, but the Azure Open AI stuff, like directly if you want to go in custom build solutions and learn about that, like what is that journey? Where do people get started? Where organizations that are interested in that? There's a lot of great resources online. Of course, the program still believe is one you have to apply for. You know, I think getting accepted into it is something that can happen fairly quickly. You're most likely going to start out with the GPT-35 Turbo model and GPT-4 access is a little bit more limited. So there may be a wait list, but there's a lot of great content on Microsoft site and there's also the Azure Open AI playground directly in Azure. So that's gonna give you a chance once you get the resource stood up to go in and start looking at having chat conversations. That's where you can sort of tinker with the temperature and different settings right to fine tune kind of the conversation. You can look at your system message which is sort of your overall general instruction for the whole chat, you know to set the tone for the bot and you can play with those. You can select pre selected ones. You can do image generation with Dolly. So you can do all of that with writing any code at all. You get right into the UI. So that's a great way to just get started and sort of see what does this look like. But then out of that too, here's somewhat recently is sort of the bring your own data idea. And so you're able to create indexes with Azure cognitive search, including even vector now. And so you can bring your own data in there, use the, what's called the retrieval augmented generation or RAG pattern to start to ask questions that you can provide on demand context from your own information so that the actual model is able to answer those questions based on your data. What's really cool about that is is you're not actually fine tuning the model. You're just giving it a snapshot of the context it needs. I like to say it's almost like giving it the Google search results and saying, can you find the answer based on this, right? The best results, you know from maybe more of a traditional search. The vector is great, but even keyword searches type work well. And so there's a lot of power there, but it's a great way to get started. Like I said, and it doesn't have to feel daunting. You don't have to know Python to get started. You don't have to be familiar with REST APIs necessarily, but you know, if you are, there's a lot of opportunity there as far as building custom applications. Well, that's been some of the most exciting demos, the richest demos that you've been seeing around kind of Microsoft's own application of the technology with co-pilot is being able to point it at your actual working docs. I mean, one thing, and I love that there's this whole, you know, the expectation or adoption curve of like chat GPT and people's reaction to it until they, you know, this is the best thing in the world. And then what kind of garbage is this doing? And then at the end it's like, oh, now I understand how it works, how to better use it. And so there's the adoption goes up. But the whole concept of being able to look at and you know, my existing data, I mean, I'm so excited for the day when it can look in my, of course, security trim, but look in my email, look in my SharePoint and Teams content, look at my chats and all that, and then I can then start, you know, asking questions and getting its help based on that, the real time, you know, learning versus data and chat GPT that was trained in 2021. Absolutely, yeah. And that 2021 dates key, you know, I think when I first got started my biggest question was, when are they going to retrain it, right? But the reality is, is even if they did, you know, as soon as they retrain it, it's out of date, right? Something else comes up. And so this idea of being able to plug in data, you know, real time, I think is really powerful and it creates a lot of great opportunities when you look at saving time for individuals at companies and really just getting people what they're looking for, you know, and I think taking things off of people's plates instead of being asked, you know, what does this particular policy look like? Who should I contact having to go through that process, you know, everyone working in different time zones, you know, waiting two days to get an answer, the ability to have a bot that can talk to you just like a person and get you that answer quickly. It's fabulous. Yeah, there's so many interesting applications. I mean, I've been, you know, as long as I've been an MVP, even back when I was a Microsoft employee, one of my agitations was task management and Microsoft not taking task management across the various applications seriously. I came from that background, project management, building, you know, project portfolio management systems. That's how kind of found my way into SharePoint back in 2004, 2005 was actually doing a project server deployment, which never worked, never fulfilled the promise, the reporting, the analytics side of it. Back then, I don't know what it's doing now, I'm not gonna, you know, talk about that, but that kind of found my way into SharePoint, but so much of that was, you know, don't tell me, organizations asking, the leadership teams might, for the, where I was consulting, companies were asking, it's like don't give me the information that I know to go and look for, tell me, you know, what I'm not asking for based on my request. So it's, again, it's kind of the core of information management problems is the search experience, is the information discovery process. Yeah. And we're really just moving into that next realm of the search experience. Yeah, absolutely. And I think that's really exciting to think about two, three years from now, maybe sooner, you know, how would the traditional Google type search, right? Taking that from a page of results again, you know, looking, clicking on the first link you think is right, scanning the article, seeing if you can find the answer you're looking for to just asking direct questions and getting direct answers. You know, once we get to that point, it's just a huge game changer for productivity across, I mean, personally and professionally. Yeah. Well, and two, I mean, again, one of the coolest demos is when you look at, like in Excel, in Copilot and Excel or just having the AI capability to be able to go in and tell, say, well, looking at this massive spreadsheet, looking at this data, you know, piece of data, what do I need to know about this? What are the trends that you're seeing? What are the three relevant things I need to be looking at with this chart and for it then highlight that? Yeah, absolutely. Yeah, pattern recognition in our complex confusing data is a huge thing. Now, I would love to be able to, instead of training the model, just be able to point it at a website and have it go and do a read and then based on the information it's been trained on what it's able to go and scan from that site, that would be a nice little add-on. And I guess you get some of that with Bing and with Google's new AI capabilities, but well, very cool. Well, so what are your other community activities? Again, you mentioned participation with like the user group stuff. Were you involved? Are you an organizer of some of those things? I'm not an organizer yet, although I would say anyone listening that needs help, please feel free to give me a call. Organizing user groups, moderation on forums, I want to continue to give back to the community. So I've spoken recently at the Kansas City user group, the Boston user group, mentioned collab days. And then as far as forums, love the Power Platform Community Forums from Microsoft and regular visitor on the Power Virtual Agents Forum as well and Power Automate. And so spend quite a bit of time there and then my blog madjimison.com and also post those on Twitter. And so that's really where a big focus has been for me, but I'm looking to expand honestly and see how much more can I help? That's always my last question is where people find you, but you just outlined it. Of course, we'll have the links. If you want to get in touch with Matt, you can find him out on, we'll have his links out on the blog and it'll be on the podcast as well as out on YouTube. So, well, Matt, I am looking forward to seeing you. I will probably see you in Boston here in a few weeks. That sounds great. Looking forward to it. Looking forward to it as well. We'll talk to you soon. All right, thanks, Christian. Bye.