AI Agents Are Here - Are You Ready to Manage Them?

00:00:06 Mark Smith
Welcome to the MVP show. My intention is that you listen to the stories of these MVP guests and are inspired to become an MVP and bring value to the world through your skills.
 If you have not checked it out already, I do a YouTube series called how to become an MVP. The link is in the show notes. With that, let's get on with the show.

00:00:35 Mark Smith
Today, we're heading to the US to discuss the future of work. Please welcome Stephanie, Modern workplace practice lead at Avanade, Stephanie's North Star is helping organizations unlock the full potential of Microsoft 365 by blending AI governance and human centred design. She's been on the frontlines of enterprise. No pilot rollouts and she's got strong opinions on what responsible automation really looks like. Full links are in the show notes. Hi, Stephanie. Welcome.

00:01:04 Stephanie Donahue
Hi, thanks for having me.

00:01:07 Mark Smith
Good to have you on. I think you've got so much to share based on what I've read on you. Yes, I'm excited to have a chat, food, family and fun. What do they mean to you? When you're not doing your tech job.

00:01:19 Stephanie Donahue
Oh my gosh, do do I pick one of each or or is it more of a general question? Yeah.

00:01:23 Mark Smith
One of each guest.

00:01:24 Stephanie Donahue
One of these. Oh, gosh, OK. So food, I think seafood and and more specifically, I love seafood. When I go to the beach. So somewhere along the ocean. Family. I have two boys, one headed back to university here in a couple of weeks and the other is a senior in high school, so it's almost football season. Looking forward to that and let's see fun. I spend a lot of time at Lake Erie, which is here in Ohio. A couple hours from my house and I'm a boater so you will definitely find me out in the sunshine, especially when we have this nice hot weather like we have here in the summer so. That's a little bit about me.

00:02:11 Mark Smith
Yeah, I I identify with the definitely the seafood. I've got 8 beaches around me, so I'm really local, you know and and where I live but.

00:02:19 Stephanie Donahue
Ohh I'm jealous.

00:02:24 Mark Smith
I'm interested in discussing your journey for the last three years particularly. And well, really two years at Microsoft have had in Microsoft 365 Copilot in Mark it end. If you were to sum up the last three years, how would you sum them up?

00:02:42 Stephanie Donahue
I'd probably call it a whirlwind. I think, you know, everybody knows this, right? But AI kind. Of. Took everyone by storm, but I'm grateful for it. I think we were all kind of trying to pretend that mesh was the next big thing, right? Or like, trying to really get into it. And we were all trying so hard. Like this is great. This is interesting. This is different than new that I think there was a little bit of. Relief when AI hit the scene. And UM, you know, it just seemed like the right thing at the right time. We were all ready for an injection of, like, what does the future look like and how do we do the future differently? And how do we innovate differently? And I think I think that's really what's driven the last couple of years. And it's it's kind of had a few iterations, right. Kind of started with just foundation like people were playing around with ChatGPT and that kind of led to copilot and then 365 copilot. Right. I'm a Microsoft focused person, so that's. I'll go but, but then now we're in the agentic AI and I feel like that hit just as fast as Gen. AI as a whole. Right. And so I think everyone was just starting to get their arms around what it meant to have Gen. AI in our lives, and now we're going agentic. And now it's like, what's this agent thing? And and now Microsoft talking about. An agent workforce and the the frontier firm and the future rate of having hybrid agentic workers working alongside of you. And I think the speed at which all of this has. To the forefront is just mind blowing right and and as. A person who's. Been in it for 25 years. It's so exciting, right? Like, I think we were all kind. Of like you know the. Cloud it's great. We kind of had our vibe. We're doing our thing. But you know, everybody had moved to the cloud and we're like now what this is the what? And I'm, I'm just really excited that I get to play in this space and. And go do this every single day.

00:04:39 Mark Smith
You have thrown out a bunch of terms there that a lot of people might not be familiar with, and so let's just unpack a few of them. One is, what's our frontier organization?

00:04:51 Stephanie Donahue
Yeah. So Microsoft came out with this concept, and there's blogs and stuff. Maybe we could share in in your notes. But the idea is that a frontier firm is the one that looks at AI first, that rethinks the way the organization works and and, you know, I think we're guilty. And I. See of just throwing tech. At people problems and we don't really rethink what are we actually trying to accomplish. And I think now this has given us an opportunity to do that. The frontier firm takes a step back and says how do we think about this differently? How do we work differently? What are we actually trying to solve? We're not just throwing tech. The you know the making, making our existing business process work better, we actually need to take a step back and say what is the business process, what are the things that we need to accomplish and how do we do that. And the idea of a frontier firm is also. So that we would have this hybrid workforce that you know, we and some of us do this already, right, if whether you have ChatGPT for your personal life or you have copilot at work, you're kind of using it in place of a person, right? Maybe for your personal life, you're asking it for recipes instead of. Calling a friend or calling your mom or whatever, right or or just going to the Internet and looking up recipes. But you're asking a question, getting an answer back, and we're doing that for advice. How do I handle responding to this person? Right. And so we're, we're now going to. This application to ask questions to get advice to go get things done, and it's replacing people in some cases. And when we look at that from. A work perspective. You know, it's a little bit intimidating because we think, well, will they? I take my job right? Because I'm now asking this application something about work. And it's giving me an answer back. And do I need to go to, you know Bob and accounting or sue and finance or you know? Bob in in marketing or whatever it is right? Like I can get answers back and not have to talk to a person, but the reality is we're all overwhelmed with too much to do. And So what I'm hoping is we're all still employed, but we actually get a lunch break and an 8 to 5 day, right. Hopefully that's not too much to ask, but the idea here. Is that this tool becomes part of your workforce. It's like a teammate. It's someone you can rely on. It's someone who can go do tests for you and that we're going to have a bunch of these, not just one, but different people to do different tasks, right? There's different agents to go do different tasks on our behalf. So the frontier firm is really about rethinking business process and figuring out how do we utilize AI to work differently, more efficiently and and like I said, hopefully someday I get a lunch break again I look. Forward to that.

00:07:45 Mark Smith
I like that I want to spotlight on on. One of the things. That you've you've talked about there, which is this? The concept of. A digital work. Yeah. Somebody that's alongside you that you know, supports you and and you talked about research functions there that you're really handing off, which would be a human in the past, which could be technology. Now one of the one of the challenges individuals have and and often people. End up in management. In that there are a good individual contributor they raise up through the org. They become management and nobody train them for management. Nobody train them. How to run a team, nobody trained them explicitly. How to delegate. And I feel that as we move into this agentic scenario here, we've got to become really good at delegating and delegating is so much more than just saying. I need you to do something because you need to give parameters. Like what does good look like? What do I expect? What's the outcomes I'm after? I don't want outputs and I think this is a big change that I'm seeing in the workforce is I don't want output I want. And it's a subtle difference, but it's a massive difference. And so how does that open up from you and me in our daily jobs, the need to learn? Perhaps new? Managerial skills, where the the thing you're you're managing is actually going to be a digital counterpart rather than. Person.

00:09:09 Stephanie Donahue
Yeah. And I think we almost make this part more intimidating that it has to be because we think, well, I don't know how to use this tool, right. And in our experience in the past with technology has been if I try something and it doesn't work, I quit because obviously just doesn't work that way, right. And I think our experience with AI is very different because we have this ability to converse. We can continue to engage with it to try and get closer to what we want. Right. And and that's part of the delegation process is. You know you you give it a command, but then you continue. If you don't get what you want back then you continue to ask it more questions, right. And we do this every day. We do this with people. You go into the doctor's office and you don't get the answer you want from your doctor and you say, but what about this? What about that right? And and we do this very naturally in our personal lives. And so. It's funny to me when we get into a work situation that we just kind of quit or just assume. Like the tool, just. Doesn't work or you know, like I even did it, you know? And I'm. I've been in IT a long time and I feel. Like I do a good job with technology and one of my earliest experiences with copilot was talking with Microsoft and giving them feedback, and I said, you know, it's great, but I wish it gave it back to me in bullet point format instead of a paragraph and just coming back to me was like, just tell it to give it to you in bullet points and. I was like, oh, right. I can do that, right? So you kind of have to rethink and be, like, not be afraid of it and just tell it what you need, just like you would if you were sitting across from a peer. And from that perspective, I think we just simplify things and and it becomes way less intimidating when we think about it as just another person. You know, some people are better than others are delegating in general. So I I do think it's a habit, this is this is not a tool that's natural to any of us as I just described. And so it takes purposeful. Effort to learn how to use it, how to work it into your work day. And I say that it's a lot like exercising, right, just exercising feels weird and awkward at 1st, and you're just like, you know, you got to make an effort to get out of bed in the morning and go workout before your day starts. But once you do it. Over time, every single day. Several months down the road, it's just part of your day. It's just what you do and you, you have this really healthy lifestyle. You don't even have to think about it anymore. It's just part of what you do. And and I look at AI very much the same way. You just gotta practice and be purposeful with it and work it into your day. And eventually I think it all of that is going to. Get easier for all of us.

00:11:42 Mark Smith
Where is responsible AI into the purview for you and and I I noticed that sound bite of responsible automation. How do you talk about this?

00:11:54 Stephanie Donahue
So I'm an information architect from Ryback, right? Like I've done SharePoint for a long time and we've been trying to have that governance conversation for a really long time and that that's part of the picture, right? It's not all of the picture, but that's what comes to mind for me first is. That initial pass of like, hey, we're just going to go light up AI inside of our organization and you really have to be careful with that, right? Not all tools are the same. Not all tools are safe and and some tools are going to surface everything you have and everything you, every document, everything. And and if you've not done your due diligence in your environment. Then you know you're at risk for, for really sensitive content, intellectual property getting surfaced or. Intellectual property being sent to to a public AI tool like a ChatGPT. That's not enterprise account, right? So we have a lot of factors here to kind of think about. We've we've got to be responsible with it and and and too we have to protect the content that we own and the content that gets exposed. To that AI and and there's, there's kind of a lot of. Layers to that.

00:13:02 Mark Smith
And so in the customers that you've worked on over the last couple of years. What are you seeing as their biggest challenge to adopting M365 copilot? And you know, I've heard things like our data is not too good security. You know, one one example I had was it's going to take us a year to get the security like we know our data is exposed, but we, you know, we use a model of. Security by obscurity. Nobody knows how to. Find it but. That's alright for us. And so they'll push back from my tears at hey, there be Dragons there. This is risky for us to go down there because we know our data estates is a shambles. So is it data? Is it more? You know I love Kawana. I don't know if you know her. And the, you know, team. And she brought up a really interesting comment at MVP Summit, which was. You know when teams came out and became the superstar tool that it did in the pandemic, nobody asked the question. His team's gonna take my job. Where we have a a, a, a cognitive loading. Now with people going is AI going to take my job and so there's resistance, right? And resistance needs normally an adoption process to overcome to take people on a journey to take them on a narrative arc to get them. From where they are. To seeing a future that they they can understand that isn't going to leave them without a job going back to what's the biggest. Challenge you're seeing when it comes to, you know, enterprise rollout of M365 copilot.

00:14:44 Stephanie Donahue
I think it's actually IT risk and legal. They're the blockers, IIT risk and legal, they're the blockers. We know for fact when, let's say you do a proof of concept internally and you only do that proof of concept with those three groups and and the point is here, they're trying to break it, right. They're going to try to break it. They're going to try to find stuff they're going to try.

00:14:49 Mark Smith
I see I see risk and legal.

00:15:10 Stephanie Donahue
To find reasons. That you don't want to implement this tool. And what we like to do is balance that because if if that's where you start, that's where you're going to be stuck because it's really hard to talk him out of it, right? You you can find something wrong with any tool if we put Microsoft Excel through the ringer like we do our tools now, it wouldn't for make it into production because of all the things that it can do. Right. But when we sit down and and look at this in parallel with the return on investment and the business use case. This then we can balance what IT and risk are saying by saying what's the risk, what's the reward and we can move things forward in that space because what ends up happening and this is something that I see risk illegal, they can, they can shut down everything all they want and think that they're protecting their business. What we actually see? On the flip side is shadow AI. If you're not familiar with the term shadow at right for those listening like shadow, it is the idea that you work around your businesses limitations to go get stuff done right. So back when we were migrating to OneDrive, we saw people who were using box, you know, Dropbox and Box. Because terrible and they were facing limitations on the size of the documents they could send through exchange. And so they go out and get a. It was. New. A new tool that would help them get their job done, and IT would have no idea we are experiencing the same thing now with AI, people are taking their business data and go into ChatGPT because they know it's going to help them get their job done. Buster and here risk and legal is like we're doing our job. We're blocking everything. You're not going to use, you know, copilot or whatever it is you want to use in your enterprise. And meanwhile, people are working around them. So. So. The people aren't. The problem the people want this, the. Employees want this. They they use it in their personal lives. They desperately want to use it at work and the biggest blockers we see are the IT risk and legal folks that just want to shut this down. And it's all about, and this is the other part that's hard about this, it's about keeping up with the latest and greatest and being confident in the tools that you're using. Do you trust Microsoft? To to protect you, do you, you know, or who? Whoever your AI provider is. Right. Everybody's got a version of it. Do you trust them? Because we've got to go on this journey together. And that's really hard to to put your career on the line and say we're going to go do this. But on the flip side, you're also putting your career on the line. If you don't because your business is going to fall behind the people that do use it, it's a balancing act, right? And I think all of that's really hard because we're moving at the speed of light here and we've all got to make choices and and deal with risk. That's definitely part of the game right now and it's it's kind of scary, right? Like it's it's a, it's a pretty big deal.

00:18:04 Mark Smith
If you could give me one C-Suite roll that was on board with getting copilot out in the organization, just one. What would it be? What comes to mind?

00:18:19 Stephanie Donahue
Our innovation people see you do see what's interesting is like the CIO's and CFO's and and you know what's interesting is that we used to look at them. As Like. The the financial people or the people who wanted to slow things down, the ones who wanted to regulate things and actually our businesses. That did the best job of securing themselves upfront of regulating themselves up front or running right now, because everything's in line and they're ready to. Go and I I do think. Sometimes we see. This, you know, the CIO's move ahead of the CTO's a little bit. I think our our technology officers are sometimes they're a little bit reserved on this. So it's very much. I think it's a mixed bag though it really depends on the company. It really depends on that sea levels, pension for risk. Right versus reward, I it's so hard to nail down one right. But I I think what's interesting is we see a lot of movement in the finance space. CFO's, CIO's, I think we we just see a lot. There, there's a lot. Of support.

00:19:27 Mark Smith
Thank you. You you confirmed my. My what I'm seeing as well, so that's that's good. There's there's there's parallels here. When it comes to adoption. What are you seeing done right as an in your clients and stuff that really, you know accelerate? UM. The adoption across the organization, you know, recently I spoke with the lead for Microsoft's internal adoption program where they rolled out to 8000 champions inside Microsoft. This is copilot. So customers 0 their journey and you know one of the big levers that that worked well is that when you get somebody totally. Anti the tick. See the light and then they. Are like because everyone's heard them being the naysayer they're all have become all of a sudden became the Oh my gosh, and they can of course, because they've been kind of vocal naysaying. They become a very much a maven and a light and A and a A, you know, a leader within the organization or a champion to. To spark growth. But are there any parallels or patterns that you're seeing that? Head an organization to success sooner rather than being drawn out over a protracted period of time, which I see where people just turn the licenses on and say good luck. There it is, but we're not doing anything else.

00:20:53 Stephanie Donahue
Yeah. And I it's almost predictable, right, when we saw this fairly early on too, where you'd have leadership come out on social media and say we tried it, it doesn't work. We don't believe in it. Right. And and I guarantee you what they did was a traditional IT. Well, they provided a training class. They gave them an hour and they sent them on their way. That's the that's the quickest way to having a failed deployment. It doesn't matter what platform, right it all of them. You, you can't do this the way we've always done IT. You know, we've got to look at it is change enablement. And Oh my goodness, we hate paying for change enablement in it. We hate it. I don't know why it is the best thing you can do for adoption in your organization. And when I say change enablement, I mean more than a training class. Do you need a training class? Sure. Yeah. That's where you can start, right. Training. But you need a community of people, to your point, like the Champions you mentioned, right? People that are trying things that they have a space to go ask questions and hopefully that's not waiting in an IT. You right, we don't want them submitting help desk tickets for help with this. We want them going to a community in, let's say, be engage Microsoft Teams. Who else has? The license who might be able to. Answer their question.  And you know, and then who's going to help them through? Like, I don't know, office hours, for example, is another great way of doing change, enablement. Help them learn from each other. Early on, I would just hop on office hours just to see what other people were asking. I didn't have any questions. I want to know what they're talking about. And then I would learn all kinds of stuff still do. Well, honestly, right. I've been using this tool for over 2 years. And I I still people blow my mind with the stuff that they come up with. So I think you have to allow it to be viral in your organization. You have to allow people to learn from each other and in order for that to happen, you've got to give them spaces to communicate. And we're all so many of us are hybrid or remote. Now it's it's you're not going to talk about it over the coffee pot. In the kitchen, you're you're going to have to find purposeful ways that people will get together and have those conversations. So it's it's it's change enablement. It's going out and making sure they've got those platforms to to talk to each other.

00:23:10 Mark Smith
I love it. I love it. Well, we're at time. It's been interesting. I've got another million things to ask you. If you put your crystal ball. I got out your crystal ball and look at the next just 12 months. So mid next 2026. Where are we going to be or where would you like to see us be in that time frame? And I.

00:23:30 Stephanie Donahue
I want to see us a little moving a little closer to what Microsoft says, that frontier firm is where maybe we're managing, creating and managing some of our own agents. We've got this really fantastic tool right in front of us. And I think today we're very cautiously creeping into things that are kind of, you know. Proof of concepts and we need to get those proof of concepts into production and we need to do it at a massive scale, right? And I know that it is going to look at it. I know the business is going to look at these, you know, big ROI changes. I think that will happen. But what I'm excited about is the innovation that's going to happen and the individual user sitting at the desk. Every day doing their day job and innovating and coming up with these really fantastic agents that are gonna. Just change the way business is done and it's not even going to come from IT, right? Like it's just going to come from that the average everyday user and and that's what I'm excited about. That's what I look forward to because we move forward and power in numbers and and all of us collaborating and doing this together is really it's. It's going to be really powerful and I think that's where change happens. If we're going to sit around and wait for formal IT projects to happen, it's going to be years. I think. I think we're all going to move faster from that individual effort and and I'm really excited about that. Yeah.

00:24:53 Mark Smith
Awesome. Stephanie, thanks for coming on The Show.

00:24:56 Stephanie Donahue
Thank you for having me. It's been fun.

00:25:04 Mark Smith
Hey, thanks for listening. I'm your host business application MVP, Mark Smith, otherwise known as the nz365guy. If you like the show and want to be a supporter, check out buy me a coffee.com.nz365guy. Thanks again and see you next time.