The Digital Project Manager

Adapting Project Methodologies for AI Without Losing Control

• Galen Low

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 28:56

Do project management methodologies still matter when AI can generate plans, notes, and schedules in seconds? In this episode, Galen sits down with Stan Yanakiev to unpack why the answer is still a firm yes—but with an important caveat: methodologies have to adapt. Together, they explore how PMs can move faster with AI without losing control, context, or the human side of delivery.

The conversation digs into lightweight project management, human-in-the-loop AI, and why speed without structure leads to “speed wobbles.” Stan shares practical examples of using AI to reduce documentation overhead, strengthen decision-making, and free PMs up to do what they do best: lead people, align stakeholders, and navigate complexity.

Resources from this episode:

Galen Low:

Do project management methodologies even matter anymore?

Stan Yanakiev:

Yes, methodologies still absolutely matter, but they will need to adapt for AI. Even when I first prepared for my PMP exam back in 2008, PMI already understood that traditional project management method could be tailored. And with AI now in the picture, we still need to know how to best manage projects. However, our methods need to support lightweight project management approach.

Galen Low:

Why is it so important to you that our project methodologies at leverage AI are still human at their core?

Stan Yanakiev:

It's because we humans build machines to help us in our life and work and not the other way around. When we are talking about human teams and projects, it's important that people, humans stay in charge with that simple application of AI. Our role shifts from being document controllers and we are stepping into the role of strategic leaders.

Galen Low:

Just getting your take on like what the future of AI and project management looks like. Will there be new methodologies to get into ferocious debates over in five years time?

Stan Yanakiev:

Well, our society and the environment are constantly evolving. I do see the role of AI in project management as—

Galen Low:

Welcome to The Digital Project Manager podcast — the show that helps delivery leaders work smarter, deliver smoother, and lead their teams with confidence In the age of AI. I'm Galen, and every week we dive into real world strategies, emerging trends, proven frameworks, and the occasional war story from the project front lines. Whether you're steering massive transformation projects, wrangling AI workflows, or just trying to keep the chaos under control, you're in the right place. Let's get into it. Today we are talking about adapting traditional project management methodologies to support AI enhanced ways of working, specifically how to run your project at speed without getting the speed wobbles. With me today is Stan Yanakiev, Founder and Managing Director of Mindrise — a consulting firm that delivers AI automation and project management as a service for digital and AI initiatives. Stan is a certified senior IT program and project manager with 25 years of experience delivering technology and transformation initiatives across energy and cross-sector industries. His specialty is software development and enterprise applications with a focus on complex IT delivery. Experienced an agile, waterfall, and hybrid approaches, he excelled in stakeholder management, fostering collaboration across distributed teams and suppliers to implement robust best in class solutions. Stan, thanks for being with me here today.

Stan Yanakiev:

Thank you, Galen. It's great to be here and discuss how we can make AI delivery both fast and structured.

Galen Low:

I love that goal. I'm excited to dive in. I think we can go where the wind takes us today, but I did sketch out a little roadmap for us. So to start off, I wanted to just get one big question that frames it all in, and we get that out of the way, but then I'd like to unpack that, zoom out from that, and talk about three things. First, I wanted to talk about the challenges that AI presents to our established models of project management methodologies, and then I'd like to explore how to go about tailoring methodologies for teams using AI, like the steps to take and the reasons why. Lastly, I'd like to get your perspective on the future of project management methodologies at large, and what, if anything, needs to change in the way that we look at them. How does that sound to you?

Stan Yanakiev:

That sounds like a clear and practical roadmap. I like that structure. It's exactly what PMs need, a disciplined approach to handle this near fast moving AI landscape.

Galen Low:

Absolutely. I love that. I thought I'd just start us off with, you know, the one big hairy question, and I'll give it a bit of context here, project people, we love to talk about methodologies and by talk I mean like debate and maybe sometimes get into a fist fight about it. But my question is this, do project management methodologies even matter anymore? Like with PMI widely recognizing that most projects use tailored hybrid methodologies and with AI like fundamentally changing the way we work, will project management methodologies even matter In a few years time?

Stan Yanakiev:

Yes, methodologies still absolutely matter. But they will need to adapt for AI. Even when I first prepared for my PMP exam back in 2008, PMI already understood that traditional project management method could be tailored. And I had been using agile methods for a few years by then. In practice, I often apply hybrid approach just like most PMs today. And with AI now in the picture, we still need to know how to best manage projects. However, our methods need to support lightweight project management approach that doesn't hamper the speed of these new tools, but instead make the best use of them.

Galen Low:

I love that. And yeah, it's a really good point, like tailoring I guess the word, the PMBOK kind of emphasized it more in my seventh edition and I guess again in the eighth edition. But yeah, it's always been there. It's always been recognized that we are tailoring, we are adapting our approaches to deliver appropriately for what we're doing. I like the word lightweight, lightweight project management. I think that will probably resonate with a lot of my listeners who are trying to move fast and be nimble and not get bogged down with all the things. But not quite. Being sure what they can get rid of. Right. Without actually, you know, crashing the car. But I really do like that focus on speed and I think that like the PM's ability to like just keep the team working at pace is the priority right now, especially in the age of AI. Maybe that's actually a good opportunity to zoom out a little bit. You recently did an interview with my team where you talked about how you have moved towards like lighter, more adaptive methods with AI in the mix while still maintaining what I believe you called, you know, just the traditional project management methods. Could you talk to me a bit about how you began using AI to make projects move faster? Like did you find yourself changing fundamental pieces around how you run projects?

Stan Yanakiev:

My goal was simple to save time on some critical yet mundane tasks, thereby optimizing the execution of our methods, not changing the methods themselves. One core problem in project management is documentation overhead In any structured project, maintaining up-to-date project artifacts and governance takes up too much time. You have duplication everywhere. Scope feeds into the breakdown structure, status updates, reflect activities, and so on and so forth. Using AI to produce these documents or keep them in sync, allows a shift to a lighter and more adaptive delivery model because it simplifies the entire administrative function. For example, my first experiment using AI was to generate meeting minutes and action items, which saved me about two thirds of the time. This efficiency freed me up to focus on high value, human-centric parts of my job, leading the team, solving complex blockers, and aligning stakeholders.

Galen Low:

I love that. You know, it's funny you mentioned that you've been using Agile for many years. You were already using it, you know, earlier on in your project management career. I know some folks, right? They choose to interpret the principles and the values as don't do documentation, which is a misinterpretation. It says do just enough documentation, like you still need documentation. But what I like about what you said is keeping things in sync like that resonated so much.'cause I know I'm like, you know, you make a decision in a meeting or you have some kind of like action item and it changes something along the way and you've gotta go back all the way through all of these documents where you've, you know, listed it out. Communications plans, you know, scope documents, decision logs, and I like this idea that A, the note taking, absolutely game changer in project management B. Taking some of those decisions and like cascading them across different documents. I think that takes so much pain out of the job that is necessary because you know, like your career, you've been doing complex software developments in enterprises. Sectors like energy where it's like regulated, heavily regulated, you need that paper trail and it becomes so much of the job just like managing the paperwork. I love that. That's like a massive efficiency gain and I think it uses generative AI and you know, the respective LLMs in a way that's playing to their strengths. I think that's really. The thing that kind of like gets me thinking is what happens when we put those efficiency gains in place? What happens when we start moving faster? Like we're coming back to like this idea of like traditional project management methods. Like after you've sort of applied AI into your workflows and your teams are using AI, how do you then make decisions around what actually needs more reinforcement? To support the extra speed so you can avoid the speed wobbles. And then do you find yourself leaning on adaptive approaches like Agile more than, say, predictive approaches like waterfall?

Stan Yanakiev:

I'd say that the more intelligence we built into a process the more careful the process has to be planned and tested with or without AI my decisions whether to use more predictive or adaptive approaches comes down to experience with project management. Experience with project management tells me whether I can realistically plan far enough. To planning more detail or whether I need to use more adaptive methods. Then there is also experience with AI that we need to have to know when and therefore what tasks we can trust it and where it will just need more guardrails because AI can suggest. Scenarios and plans, but we have to remember that by design it's probabilistic. If you ask a LLM model 10 times the same question, you may not get 10 times the same answer. And it's our view of reality and predictions aren't guaranteed so much with what is actually happening. So it's us, the project managers who need to be the ones to keep things grounded and in check and use our professional judgment and skills to get the benefits of AI while avoiding some shortcomings this technology still has.

Galen Low:

I actually wondered if you could like, walk us through, I guess, your journey with, for example, like an AI note taker and, you know, I guess what we're talking about is, you know, being the human in the loop, right? You can't just take it at face value, expect it to give the same results every time. I guess my question is like, when you started using AI in your projects, a note taker or you know, any form, how much time were you spending just double checking to make sure it wasn't hallucinating or having the wrong idea? Has that time that you spend, you know, being the human in the loop gone down? Or is it kind of, has it been like consistent where you're like, okay, I know I need to spend five, 10 minutes just looking through and making sure it cascaded the information correctly across the documents?

Stan Yanakiev:

Well, when we first started experimenting with AI, I didn't yet have enough experience with it to know. How it works and what I should do to get the best outta it. So I did treat the outputs from AI as just a draft and reviewed it and before feeding some sensitive information, I did anonymize it. And then I replace the various the names back in the output. But it is essential to keep some processes in place, some checks in place, like follow treat output of AI as a draft. Don't fit sensitive information in open models. And basically double check everything just to avoid hallucinations and facts that are not true.

Galen Low:

I guess it's I know people say this, but you know, even as you're explaining that, it's kind of like having a human on the team, right? It's gonna rely on you training them. Well, and at first it might seem like, you know, you're just figuring it out. They're just figuring it out. Their answers might not be the best and you kind of need to set up some systems to provide feedback to review. The lovely thing about AI in its current form is it's learning as you go. It's not just a tool that does one specific thing. It is probabilistic. It's not always gonna give you the same answer, but at least you can continue training it and giving it feedback so that it gets closer to that answer most of the time. Then I think it's the trust, right? Like it's like when you don't know it and you don't trust it, you probably put more checks and balances in there, and then afterwards as you sort of begin to trust it, you know what to look for and there's still checks and balances, right? You're still anonymizing data going into it. You're still double checking things. You're still being that human in the loop. It sounds to me like actually the time you spent at the beginning figuring things out, building that trust. It's probably less now, like now that you've kind of got your flow in there. The reason why I'm asking that is because I know there's a lot of folks who listen to this podcast who are maybe a bit discouraged because this thing that's supposed to make our projects go so fast actually feels really slow because they're double checking everything. They're like, I probably could've just copied and pasted this into all seven documents myself, and it would've been faster. But actually the efficiency gains happen later as the models, as the tools like learn.

Stan Yanakiev:

It does take some time. We need to treat it as just as a tool and do our homework and learn about it to learn about prompt engineering, how it works, how to achieve best possible outputs from it. And yes if we apply it across the board we can, obtain efficiencies from it as long as we use it correctly in the most effective way.

Galen Low:

I wanted to go back to something you said earlier and see if we can apply it to AI because, you know, I started this out saying people debate over methodologies is the agile versus waterfall. You know, fist fights you had said earlier, you're like, okay, well I've got enough experience in project management to know which tool to use when. When to be adaptive, when to be more predictive based on the constraints of the project. If AI is also a tool in our toolkit or a set of tools, I guess I should say, are there some cases where you actually might decide not to use AI for a specific project?

Stan Yanakiev:

Well, probably yes, depending on the specifics of the project and whether using AI would really lead to some benefits or, maybe just taking the time to work with it will burn all these benefits. So we need to always evaluate when it is useful and to what extent and when probably not.

Galen Low:

Yeah, I think there's a lot of folks out there who feel like this is all or nothing, and if it's nothing, then they're gonna get left behind. But it's really interesting thinking about right. How we choose when to use, what types of methodologies or approaches and when we need to tailor them, right, for the constraints of our project. And then, you know, AI could be, I mean, I know there's a lot of folks out there trying to just standardize the way they use AI in projects so that every project goes faster. But I like the idea that you know, only if it's useful, only if it's appropriate for the project. You know, I've been following you for some time, and I feel like you are a very strong proponent of some of the things we're talking about, right? Human-centric project management through AI, as opposed to some folks that I've spoken to who are like actively trying to achieve fully agentic teams with like very little human oversight. I wanted to ask like, why is it so important to you that our project methodologies at leverage AI are still human at their core when AI's capabilities are already approaching, you know, parity to some human capabilities at the project level.

Stan Yanakiev:

It's because we humans build machines to help us in our lives and work and not the other way around. Agent AI is a powerful technology that is perfectly fine to use in scenarios where there's no need or of human oversight, or perhaps it's not possible due to scale, speed, and complexity. But when we are talking about human teams and projects it's important that, people, humans stay in charge because there is much more to a human than just computational intelligence. If we go back to the simple example of using AI to automate meeting minutes with that simple application of AI, our role shifts from being document controllers. We are stepping into the role of strategic leaders and we can then dedicate more time to high value human-centric activities like stakeholder alignment and proactively solve complex blockers. The whole delivery process then becomes more focused and in a way more flexible. AI has its place in that, but it's the place of a helper or assistant, not the project manager.

Galen Low:

Yeah, I really like that the sort of like human-centric parts of the role that honestly they get underplayed sometimes, but it is still about people, stakeholder expectations. It's still about negotiating. It's still about. Making really critical decisions about a business that are complex, that are, you know, based on information that, you know, maybe it would be prohibitive or just not very effective to train a model to help you with that decision. And then fundamentally, I think we're at that point where I don't think we'd let AI make a big decision for a business on its own anyway. So there's like the humans in the loop, but also the humans are just, you know, we're there. It's still human led collaboration. I really like that. I thought maybe I'd put you on the spot a bit because I totally agree with all of what you're saying. I still believe in, you know, all the rigor of project management, you know, and whether you call it traditional or otherwise, we're talking about AI and projects moving faster. We've used, you know, note taking as an example, documentation as an example. I'm wondering what your favorite traditional project management sort of tactic or method is that is still hyper relevance in the age of AI. I'm thinking things like, you know, I know there's people out there, actually people very close to me who really hope that AI will get rid of a risk register or, you know, replace a decision log or, you know, can make us not have to create a communications plan. You know, like skip documentation, but what's the one, what's your favorite traditional project management tool that you think needs to stay in place to keep us all in check as we go faster with AI?

Stan Yanakiev:

We always need to have a plan for our project, whatever format it takes, whether traditional or agile. But I would still keep my old Microsoft projects to manage my traditional projects. And I would use AI in just about every project management process to provide input to help me make decisions. So I would put no constraints, but then I would review the output and make sure that the, it's what I'm looking for. Also in terms of scheduling, I said I would use Microsoft project, but I still consult AI. Whether it can produce some schedule that or some tasks and activities that I have probably overlooked.

Galen Low:

I love that answer because there are also people in my circles who hope that AI will mean we don't need project plans. And I like your answer because. Project plans are a communication device, and then especially scheduling, right? It's like, you know, the logistics of it, but it's still about setting expectations with humans and showing people kind of what the future looks like and what's going to happen. It's actually expectations management more than it is what people would consider, you know, a sort of timeline of tasks. I think a lot of folks will be like, yeah, I don't need that anymore. AI will just tell my team what to do and when. And it skips over that idea of people getting comfortable with what's about to happen, you know, and why and how the puzzle pieces fit together. Last little random question for me. Do you put AI stuff in your risk registers? It's AI, a risk in your project?

Stan Yanakiev:

I would do that, but I, again, I would review the output from AI. I would definitely use AI to suggest some risks that I can put in the risk register.

Galen Low:

I like that. I like the sort of things that you may not have thought of, like as that kind of assistant, that kind of tool, that kind of support because you know, our projects are complex and you're across a lot of things, and yeah, you'd be forgiven and your team will be forgiven for not catching every risk in the world, right? So, I like that. Well, actually I lied. One more. Do you put AI in a RACI chart?

Stan Yanakiev:

No, because I treat it as a tool.

Galen Low:

Yeah, I thought you were gonna say that and maybe that actually takes us into the future a little bit. And I thought maybe we can round out just by like. Maybe just getting your take on like what the future of AI and project management looks like. Will there be new methodologies to get into ferocious debates over in five years time? Or will AI close some of the viability gaps of like every project team having their own sort of bespoke methodology? Will it like effectively end methodology debates forever or maybe not so drastic?

Stan Yanakiev:

Our society and the environment are constantly evolving and they will keep evolving, so I can't imagine that we will stop discussing how to best do things and our methods and approaches. But with AI now in the picture, I do believe that there will be probably less scope in this discussion to come up with some new methods and approaches. Because ironically, I think that this intelligent technology will help people in governance roles to see more clearly what doesn't work. But we will keep discussing. And many people are talking about artificial general intelligence with broad cognitive abilities. But to be really useful in project management and being in other fields, I believe that AI also needs to become more specialized and more closely tuned to specific, domains such as project management, types of organizations different ways of working so that it can help us better. And as I've already mentioned several times, I do see the role of AI in project management as a really helpful assistant team or aspects of this exciting profession, helping us manage projects more effectively and successfully. And strangely enough, liberating us to focus on what we humans do best.

Galen Low:

I like that. There's two things in there. The first thing that was my aha moment when you started in on that, you know, I was, we were talking about like, oh, debates of, you know, agile versus waterfall. But I think it's a really good point that humans are always gonna be debating what's the best way to do something, whatever it is. It's gonna be the best way to use, you know, your project planning agent. It's gonna be the best way to have an AI, you know, agentic AI team member, you know, participate in standup. We're always gonna be debating these things. And in fairness, even though a lot of the debates seem to go nowhere, they seem to just be stubborn, sort of stagnant debates. Actually the ones that are productive actually help us figure out new ways to move forward. And you know, I think that's always been the case with humans to sort of debate and challenge one another, to find a better way to isolate what's working well and what's not working, and then coming up with ideally a better approach. But then I also like what you said about AGI, because I hadn't really thought about that. Artificial general intelligence kind of goes, you know, up and broad into, you know, to be as generally intelligent as humans are, but then taking that and then going narrow again, which is what we would do. I mean, what we have done as humans too, where it's like you take that general intelligence and then use it to specialize in a domain. That's where we become even more useful. I am whatever physically capable of cutting down a small tree. Am I good at it? No. Like I'm not a specialist. I will never, you know, be that person. But I really like that idea that even when we get to AGI, most people think of, you know, dystopian sci-fi futures, you know, Terminator movies, Skynet, things like that. Just taking over everything. But doesn't have to be that. It could be general intelligence that allows it to be a better coworker, supporter assistant, and yeah, maybe I guess moving beyond tools, but still being helpful in a specific way with that domain knowledge, whether it's AGI or not, but that AGI could lead there. Very cool. Stan, thanks so much for this. Just for fun, do you have a question that you ask me?

Stan Yanakiev:

Are you planning to make a podcast with an AI project team member?

Galen Low:

Like as the guest?

Stan Yanakiev:

Yes.

Galen Low:

You know, I hadn't thought about it. It's a brilliant idea. So it's funny because like I find myself playing with Notebook LM a lot. And one thing I did was for one of my talks, I fed it into Notebook LM and I asked it to sort of challenge it, criticize it as a podcast. And so actually I really enjoyed listening to it. These two AI hosts, co-hosts ripping my talk to Shred, and I'm like, this is actually pretty viable. I hadn't thought about it for myself, but actually now that you've planted the seed, I may have to do that. Yeah, a podcast interview with an AI project team member. I really like that. If I did it, would you listen to it?

Stan Yanakiev:

Yeah, definitely.

Galen Low:

Awesome. Stan, thank you so much for spending the time with me today. This has been a lot of fun. Thank you for your insights.

Stan Yanakiev:

Thanks for having me. It's been a pleasure talking to you.

Galen Low:

Just before I let you go, if there's any folks who wanna learn more about you and what you do, where can they find out more about you?

Stan Yanakiev:

The best place to find me is on LinkedIn and you can also check on my posts and videos on our Mindrise company blog.

Galen Low:

Awesome. I'll include those links in the show notes for folks listening who want to check that out. Stan, again, thank you so much. Alright, that's it for today's episode of The Digital Project Manager Podcast. If you enjoyed this conversation, make sure to subscribe wherever you're listening. And if you want more tactical insights, case studies and playbooks, head on over to thedigitalprojectmanager.com. Until next time, thanks for listening.