Unveiling the Future of AI and Low-Code: A Deep Dive with Andrew Gaskins

Mark Smith: 0:31
Welcome to the Power Platform Show. Thanks for joining me today. I hope today's guest inspires and educates you on the possibilities of the Microsoft Power Platform. Now let's get on with the show. Today's guest is from Colorado in the US of A. He works at Lumen Technologies as the global Power Platform lead. He focus on the administration governance of the Power Platform inside the organization. It's not his first time on the show. You can find links to his bio and socials in the show notes for this episode. Welcome back to the show, Andrew. Hey Mark, thanks for having me back. Good to have you back and I really am looking forward to understanding how things have changed and what has changed and also lets to drill in more into running the Power Platform practice inside a large organization. For those that didn't hear the last episode, you can find it at podcastnz365guycom, when we unpacked what Andrew did at Lumen Technologies, what Lumen does and really how he scaled the Power Platform. Before we jump into that, is there anything new? Anything new in your private life that's happened lately that you can share, anything, any milestones reached or anything like that, Andrew?

Andrew Gaskins: 1:51
Yeah, so I just recently became a principal solution architect and mostly M365 area here at Lumen, and so that's been really exciting, really good stuff.

Mark Smith: 2:02
Yeah, yeah, yeah. Have you attended any great events since I last spoke to you, which was all coming up close to a year ago? I think we're about 10 months ago last time I had you on the show.

Andrew Gaskins: 2:14
Yeah, nothing. A lot of things online, I think the things that we've all been attending mainly build and some of those things. But looking forward to the conference coming up here in October, that's going to be. I think that's going to be really special.

Mark Smith: 2:28
So you're going to Vegas.

Andrew Gaskins: 2:29
Yes, going to Vegas. Nice, nice nice Still can't get the YouTube tickets though.

Mark Smith: 2:35
Yeah, tell me just some high level stats, because it'll be interesting to see how things have grown. How big is Jamaica Community inside Lumen?

Andrew Gaskins: 2:48
Yeah, it just crossed over the 2000 mark for our major community.

Mark Smith: 2:54
Yeah, yeah, it's really amazing.

Andrew Gaskins: 2:57
Do you?

Mark Smith: 2:57
ever feel for the number of automations, chatbots, apps that you're running now.

Andrew Gaskins: 3:04
Yeah, as we're kind of looking at it, and we've had some divestitures and the size has changed in our company and we've still maintained probably around 8,000 automations running and just over 5,000 apps that are currently running. We're seeing pretty significant growth every single month. Yeah, it's been year over year. It's really really great growth as well.

Mark Smith: 3:36
One question I wanted to ask you was around how do other parts of the business? I'll give you a scenario that I've come across. So an organization sets up a power platform practice and they want to go out and empower people in the organization to become makers and build their own solutions. Really, what's been happening for the last 30 years where people have gone and grabbed Excel and built something to make their life easier in their department? or an access database or you fill in the blank right any product. But there's pushback now that I'm starting to see that managers say, hey, this person has been employed to do this role. They're not to be an extension of the IT department. If you go find your own people for that, this person needs to stay within their lane. Have you come across that and how do you deal with it?

Andrew Gaskins: 4:25
You know so surprisingly? No, we actually haven't come across that. We, I would say we have the inverse of that, where we have people within departments who have gone and they show what they've done, where they've been asked to do a certain thing, and then either directors or managers within those departments then see what was done and they go and ask them to do more and more. So now all of a sudden we see these people who have a backlog of things that are being asked for. So it's interesting On that side we're not having that, and I think part of it is because the business is seeing the value that's invested there. So they're understanding that if that employee is over there working on something that is going to either increase revenue or bring down the amount of time for a field visit or something like that, they're seeming to grasp that really, really fast.

Mark Smith: 5:25
Yeah, interesting. What's the? You know, since we last spoke, a little technology blew up in the market called large language models and generative AI, and I suppose a lot of people are thinking what does that mean for the future of employment, work, that type of thing? You know, are we going to live in a utopian society in the future or a dystopian society? And I see that it's almost a responsibility for employers in some way, as well as employees, to really, you know, I love this term that came out of Microsoft, which is learn it all right, and really, how do you embrace? And you know, if we're going to see a future where we're going to see artificial intelligence be bought to the bare a lot more, I would be in the position to going okay, if I can see that future and I don't know what it's going to look like, but I'm going to go out and I'm going to learn everything I can about enhancing my role, my position, what I do and what I see. The future of that role is embracing this technology, learning how to use it. I'm probably going to be at an advantage to somebody that says I'm not going to bother learning it. You know, it's fearful, it's the unknown, and I just see that the power platform is such a gateway for people to really bring their digital literacy to a new level. It's got automation in there. We've got AI builders so you can start using. We've got co-pilots going to be rolling out everywhere. What's your attitude to AI and how it will potentially empower your organization in the future?

Andrew Gaskins: 7:07
Yeah, you know, we're kind of in the middle of that journey right now and we're looking at what our generative AI framework is and I would say we're moving fast, but we're also moving slow, and maybe not slow, but we're being very cautious with how we're doing it. There are a lot of ramifications with who we work with, an idea of code being generative and how that all works. What you were saying there about Satya's being learned it all, not to know it all, is, I think that's it. We were actually just having this discussion tonight with our son about, kind of, the role of generative AI and talking about what's happening with schools and things like that, and it still seems like the people who are going to be most successful with this are the people who still understand their domain and have a knowledge base. So, because you can go and you can ask Power Platform, you can use co-pilot and you can say, hey, go, create this expense reporting app that does X, y and Z. But if you don't have that domain knowledge, then you're at that edge there and I think it's that idea of either it's either the front end of build out this framework for me, so that then I can take it and I can extend it. Or you have that last mile where it's like, okay, I have everything here, or I've written this essay, for instance. Now let's take it that last mile, read through it, proof it, make it shine a little more than I could. So I think, when we look at AI from the side of it is a tool instead of a replacement, and I think it's like anything in history that we've encountered. I mean, you could say the dictionary was going to replace writers, or it's back to the car or everything like that, and I think we have. The concerns that I have are around bias that's built into the large language models, and I think those are going to be our biggest concerns, more than the fear of it's going to replace me or what I do. But I will say that book, that AI2041 that you recommended, has been probably one of the most interesting reads, because those stories sit with you for a long time afterwards and they really do make you think about implications and ramifications of what you're doing with the technology.

Mark Smith: 9:46
Yeah, yeah, totally. I saw a TED Talk recently where an educator in the US had they have shown that the higher ratio of teachers to students sorry, if a teacher has a class of a larger class, each individual students get less attention and definitely less tailored attention in that classroom environment. And they were able to show and I forget the academic institution that by creating a chat bot mentor tailored to the individual, the learning outcomes went through the roof for the individual because it was one on one tuition and the other thing is and particularly interfacing with the courses, the thinking of the individual and how they got to it was all fed into the AI model so it was able to adjust based on the person's learning styles and what they found is that if you could ultimately drop a teacher student ratio down to one to one, the learning goes through the roof. It's exponential, the learning potential, and so the academic that was presenting at this TED Talk was showing that the implication was Z, as this rolls out and becomes more prevalent. And you don't, as you use the dictionary example, you don't go from let's ban it because they're gonna write the essays or let's ban it because you know we can't detect it anymore. It's about going on a second how could you enhance the learning ability of each student so that when they're, let's say, doing a math equation, that it understands the bit that's tripping them up and then Tailors the, the situation so that the learning, actually the learning outcome, happens and the shouldn't goes to the next level, you know, and increases your ability, and I just think that's an exciting, some exciting times that we're in right off that concept of being able to you know, detect. You know, I know I was never good at math. I was well, I wasn't good at school period. It was, yeah, yep, model to my learning style, for whatever reason and I hate it. I hated it, and, but I'm thinking. Imagine a scenario where the things that I found difficult were observed way sooner and Corrective measures, measures that were tailored just to my individual learning style, was put in place. I wonder what the impact could have been for me Growing up. But then I think of the future generations. What will the impact and like potential be for, for, you know, individuals in the future?

Andrew Gaskins: 12:25
That's exactly it. I mean, if you you know, and kind of back to understanding your domain, if you understand your domain, then you understand prompt engineering and I think you know that's gonna be. You know we were talking today as well, as well, like with my son, like okay, so he's, you know, he's going into sixth grade. How is this gonna work in terms of, you know, understanding technology? He's gonna be going to a micro school where that's less of a focus and the focus there is really about, you know, communication, collaboration, those type of things, and you know we talked about those are the things that are gonna be more important because that is the language and you're able to communicate. If you can communicate, then you can think through those prompts. And yeah, I, you know I spent this afternoon, you know, working, you know, on some pilots with, with co-pilot and in, you know in word, and asking it to Give me some video scripts, you know, for how to is on different things in the power platform. I like it. You know things that are that people within our area are starting to have issues around, like you know what are security roles or you know what are teams inside a dataverse. So I was able to Through the correct prompts. You know, say, give me this video script and point it at the, the, the Microsoft learn information, and it comes back, yes, and it gives me a really amazing script with your intro, with your example and your conclusion, and and then go through and read it and you're like, well, you know, that's maybe not quite. Either that's not quite right or that's not how we do it as a company. You might have some different ways that we do it, but again, you know that. You know that saves so much time. It's, it's, you know, I think it's it's about Freeing us up to do the work. That matters. Yeah, as we go, I like it.

Mark Smith: 14:07
Yeah, in your power platform practice. What do you call it? Do you call it a practice at Lumen?

Andrew Gaskins: 14:13
We actually don't? We just kind of it's just a, it's just our team right now? And then a material, okay, okay yeah, yeah, and we've, yeah, we, we, we've absolutely stolen your Enablement. So we do, we have, you know what? We have our center of excellence and we have our center of enablement, which is which is really the makers. Yeah.

Mark Smith: 14:32
Yeah, nice. How do you onboard new individuals? You know, I forget you recommended somebody come on the show and I just the names has slipped me. Oh, mohammed, their organization. Who was it, mohammed?

Andrew Gaskins: 14:45
Yes, yes, yeah, I'm a home and I was a home.

Mark Smith: 14:48
He has this drivers license kind of mentality for newcomers yeah, they've got a, they've got to know the rules of the road, of how to operate Inside your organization, and I was like, wow, what a, what a brilliant idea. How are you onboarding new folks and how do they hear about your, your team? How do they get involved? Do you curate that, or is it up to them to find find that, to find their way?

Andrew Gaskins: 15:16
Yes, I would say that we're moving into phase two now of adoption and, honestly, it's so often like listening to the guests on your show, it's like you've got to have a notebook with you and Mohammed's interview was a perfect example of that where you're like, oh, I did not even think about that. You know, you kind of start to take these and pull them together, and I think what we're finding too is that, as our, as our citizen devs who then become makers are becoming more adept at the technology, that we're also now having to shift how we do things as well, because the demand is getting greater and the types of projects are becoming more technical. And I think we talked about this last time about kind of like the fallacy of, like the low code, no code. But I think we're seeing, we're seeing now that that these makers are understanding more and more every day. And so what we're looking at is you know, we have that kind of initial idea of you know, we have the default environment and that's, you know, it's where everybody landed at first, and then, with our license, we looked at our environment strategy. We looked at you know what's our, what's our maturity model and how are we sitting with that. And then from there we now we are looking at dev environments and, you know, trying to do the routing to the dev environments for new users and you know a lot of the we're using a lot of the managed environment features now as well, excellent, so yeah, so, like the notification the first time you come in and giving information. But where we're trying to move now is to say, okay, look, if what you're doing is personal productivity, we're going to assume that you're kind of just tinkering and you're going to figure out. The things that you want to do might be you know, the notify me when my boss sends me an email, type of stuff that we, that we all started with. And then we're going to start seeing people who say, well, hey, I, I want to use Power Automate to do X, y and Z, and you know it's going to have an impact for my area. And we have discussions with them, first about the licensing for that. You know an understanding there. But then where we want to go is to say, all right, if you are looking to take that next step, for instance, so we're going to go from you being a citizen developer to what we would think of as a maker, then we want to set you up for success by giving you, like, baseline training. So Intro to Power Platform, power Apps, power Automate, all of those the what. The idea there is that then we lead them up to App in a Day, because what we found was people would come into App in a Day, but if you come into that very first, there's a certain point in it, no matter who you are, you're like a deer in the headlights. So we wanted to give them that baseline. That's only like three hours. You know, three hours of work to get to that. And then the other thing we do is we take that App in a Day and we split it into two days. So we do half a day each day, which helps, which helps people within the community, because they don't have to take an entire day off. They can take the morning, yes, you know. And so we ate 12,. You do that, and then you have a little time to digest it. You have time to either go back or finish things, if you can, and you know, then at that point we're going to say, okay, so we're kind of considering you to be at a maker level now. And then the next step would be how are you going to start deploying what you're doing and who are you deploying at four? And then that kind of gives us into the secondary of saying, okay, well, you're your HR and you've got these 10 apps and we know that that should probably actually be in dedicated environments. So we're going to, we're going to give you your own stack and we're going to give you your dev test and your production. So we've got that gateway. We're going to get you set up with pipelines, so power pipelines, so that you can actually, you know, push things through properly. But as part of that, what we want to start doing now is looking and say, okay, we want you to to, you know, test, to taking our you know, our data versus environmental fundamentals, our solution training, you know, which is going to talk about security roles, environment variables, things like that, and then also understand pipelines and, ultimately, and GitHub with that, so that we can, you know, get a lot of people to understand what's going on. And then we're going to get our data back to you and we're going to get to the main thing that we're going to do with that, so that we can kind of get that whole thing going. Once they've gone through that, then we'll grant them access to that stack, nice, and some of it might be that you might already have the, the leader, and that's something we're working on right now too and something that you blatantly stole from Alan Chai. You know, with the way that they do their kind of pods, I think we work directly with them and then they work directly with their community. So excellent, so that person is really going to be the one responsible for understanding all that training the other people within there may not have to have, like the pipelines in the GitHub, because the person who's the, the, the headmaker for HR or for that pod is going to be, is going to act like the release manager and we'll handle some more of those technical side of things. So so that's kind of you know, that's kind of how we're doing that. And then you know, then there's the next step of trying to roll in the you know the fusion and pro devs on this and you know, getting them into the you know specific app in a day is like the low code for developers and the, the PCF power apps code components app in a day. So those are kind of that's kind of how we're looking to grow this over the last part of this year. And then we have kind of our, like our monthly events, which are the, you know, the Lumen makers were were because we're seeing how our community is growing in terms of their, their topics and their understanding. We're going to introduce brown bags on top of the Lumen makers once a month, and then probably an advanced topics as well, and then we hold office hours, yeah, eight times a week. So it's going to be a big undertaking, but we feel like at this point it's really important that it's done this way.

Mark Smith: 21:28
Yeah, let's just jump to what you talked about.

Andrew Gaskins: 21:30
That was a long-winded answer.

Mark Smith: 21:32
That was great. I just want to jump to what you said there towards the end, which was around you've taken care of the makers and then you've got your fusion and your developers, and I think of it a very similar way and I'm interested in the developers in that I find that developers in my experience, not all but some are opposed to the power platform. You know, if they're used to developing their own solutions until they're not, until the light bulb comes on and they realize that security, plumbing, scaling, all that kind of stuff it's taken care of, you can just get to business value super quickly. Alms, salauders, you know, supported pipelines, all those great, rich things that have been developed over years is all taken care of. And now what you can do is focus on building value, really business value to the business, but very rapidly, without putting the plumbing, the infrastructure, all that kind of stuff that has to be in place, the governance, the risk of, you know, legacy code and then technical debt associated with those. Every environment, you know, every solution that anyone builds needs that in place. How do you help your dev, your hardcore dev folks, go from that to that light bulb moment and then becoming a real strategic part of your team, because there's so many things that they can build and functionality they can provide to your fusion makers and even your, your makers that can, really to do things that they could make a you know, wouldn't know how to do, whether it's a PCF, control components, any kind of thing like that. That's kind of connected outside of their sphere. How are you helping your hardcore developers understand the power of platforms and ecosystems?

Andrew Gaskins: 23:26
Yeah. So I think part of that we're seeing a natural curiosity with that and I think we're the kind of the gateway to that is usually the custom connectors and seeing how quickly they can work with that, what. What we're planning to do, as with kind of this whole rollout and to to gauge the interest by the developers, is not to so much maybe try and go push it on them or tell them they should use it, but within, within, like our PMO, when you know they're looking at starting a project, making sure that everyone has an understanding of the tools that are there. So when, when they're ID80 and they're trying to look for what's the right solution to this, they they won't see it as a as a competition for what they're doing currently, but instead to look at it and say, well, to your point, you know, everything is baked into this, 80% of what I have to do and I can spend 20% of the time and get this over the line, which then gets me on to the next thing, which may not necessarily be the right thing for the power platform. So I think just getting them to view it as a tool and not not a competition for what they do Exactly that's, that's how we're kind of hoping that they'll go about it, and I think you know thinking like, look at, like the GitHub co-pilot and things like that. I think the you know embracing that mentality of, okay, this can free me up to do the hard cool work, and so that's kind of that's what we're hoping will be the selling point on that as we, as we move forward.

Mark Smith: 24:57
Yeah, I've been. I've been referring to the power platform as as gateway to AI for a lot of organizations lately, and in that there are there are companies that are like Adir and headlights when it comes to what are they going to do around AI. They know they can't use public AI's and so therefore, there's the whole thing around. How do you use things like OpenAI and Azure that you make sure it's only got access to your data sets? What would happen for your organization if we painted a picture of all your data was accessible, every data set in your entire organization?

Andrew Gaskins: 25:32
So in other words, it had the right governance permissions.

Mark Smith: 25:35
It was cataloged. Yeah, do we have a future state where you know because I'm just thinking in my organization if I had an AI that was aware of everything, every training resource. My company's been going 111 years, right? You imagine all the data sets that must exist somewhere right.

Andrew Gaskins: 25:56
Yeah.

Mark Smith: 25:56
But it's, it's severed. You know there's no access to it. There's no for what, there's been no need, if you like. But in a world where, if you want to have as we, as you mentioned before AI that's unbiased, etc. Generally it's going to be on your own organizational data which goes back to then, is our data biased and you know what do we need to do, and I talk about cataloging and getting that clean and consumable. But what happens when you get to that scenario where all data that within the organization is available is consumable in a governed fashion and we've taken care of things like compliance on that?

Andrew Gaskins: 26:39
data.

Mark Smith: 26:40
Do we get to a world where, within five years, that there is no such thing as low code? No code, it's just no code. In other words, you, just you. The the prompt engineering is I need a solution that does x and it creates it for us. I can only see it possible if that the data is all available. I think the data is the most critical element.

Andrew Gaskins: 27:02
Yeah, I think so too, and I think that's one of the biggest challenges for probably every organization is is properly cataloging and tagging your data to make sure that a the access is is there for the people who should have it. But then, yeah, that idea, you know, I mean, I think we went back this years ago thinking that you know, if we, if we had access to all of our excel files, to the SharePoint search, what would that do for you know, what would that do for the company, if you could see all that data at one time, which never really truly realized itself the way that it was, kind of it was put out there. But I do think once we do that and you start to see that and you can, you can look at those trends and you can see the areas that that you have either missed or are missing, because it's just not something that that we can see, then I think and it goes back to again it's less again about the technology and back to the understanding, your domain you know. So this it just kind of gets back to the point where, so then maybe someone doesn't have to go spend 40 hours. You know a week building. You know building apps like they are right now, but they're going to have that knowledge to get them to that point, because I don't think that you're not going to be able to ask AI to do it all for you. There's always going to be tailoring for every situation, so, again, you're just going to be able to, I think, work smarter and work faster with that data behind you. I mean, I think there's going to be, I think there's going to be broad bumps with it. I don't. I think the next few years are going to be really, really interesting and I think we've got a significant responsibility to make sure that those of us who are working in it are are treating it with the seriousness that it should be and understanding the ramifications for it. You know, not only at a legal level, but you know at a personal level, or you know interpersonal level, those type of things I like it. I know it turns almost up and it's like you know I'm not going to be able to do that, but you know at a personal level, or you know, interpersonal level, those type of things I like it.

Mark Smith: 29:02
I know it turns almost up and my last question for you is what do you, what? What excites you about the next 12 to 24 months?

Andrew Gaskins: 29:09
Yeah, what one thing there's. You know the there's just so much with what's happening right now. I think what I'm most excited about is to see our maker community grow when they have access to the tools. You know once, once we feel comfortable as a company, to bring for Power Apps and Power Automate and get our community to the point where they understand again that prompt engineering, how to ask the questions, how to fine tune, how to make their code better, what they are writing and what they're doing. And I'm excited to see the kind of the baseline right now of how our apps are done, what they're doing, to what they will look like in 12 months, and I'm sure it will be significant. And I'm also excited to see how much like, like, how much of the like Excel data will consume and bring into dataverse because of how easy that will be. So that's going to be where we really start to see it shine and like again, that's just, that's just the tip of the iceberg. It's such a and it's such an easy thing. Really, that's going to be. It's going to be pretty cool.

Mark Smith: 30:15
Awesome, Andrew. Thanks for coming on the show. Thanks for having me, Mark.

Andrew Gaskins: 30:18
This has been a great, great talk.

Mark Smith: 30:20
Hey, thanks for listening. I'm your host business application MVP Mark Smith, otherwise known as the NZ365 guy. If there's a guest you'd like to see on the show, please message me on LinkedIn. If you want to be a supporter of the show, please check out bymeocoffeecom. Ford slash NZ365 guy. Stay safe out there and shoot for the stars.