Mind the AI Gap — Value Streams and the Adoption Curve

Hey everyone, welcome back here to textrink TV. We've been talking a lot about the upcoming Flowtopia conference on June 24th, and we've been highlighting some of the speakers and sessions, as well as what's going on in the world of value stream, value stream networking, mapping, et cetera. Our next guest is a presenter at Flowtopia.

His name is Brian Penisha. I hope I got that right. Hey, Brian, how are you?

Welcome to textrink TV. Great, Alan. Wonderful to be here and love talking, very passionate about value stream mapping and value stream management.

Very cool. I should mention Brian is the CPO at Enverus, and we're going to find out a little bit about that in a second. But I wanted to take a little time, Brian, and give people a sense of who you are and how you came to be in Enverus and why you're talking at Flowtopia.

Yeah. No, great. Thanks, Alan.

So, I think my journey, I started in software development and architecture. I'd say about the first 10 plus years of my career was in healthcare. And I think healthcare and understanding the systems thinking view of healthcare systems in particular.

When I was early in my career, I was a software engineer, I was an architect, and when I was building these systems, I was shocked. I wouldn't say shocked. I was astonished at how much complexity, layers of management, kind of like how the left hand doesn't know what the right hand's doing, and it became very apparent to me that bureaucratic challenge was actually costing lives.

And then fast-forward, I ended up taking care of my grandfather, a member of that greatest generation. He had a battle at late stage in his life, in his 80s. And I at the time was an architect at Johns Hopkins, and even with my inside knowledge and awareness of the healthcare system, it was like a full-time job navigating the bureaucracy to get him the care he needed.

And I started really thinking about that as I moved into technology and working for companies like Capgemini and General Electric and traveling all over the world and using architecture as a way to streamline and make more efficient the world that we were modeling. It always came back to me that the missing piece was that the very human dimension of how we're organized, and that led me to build companies. Enverus isn't my first rodeo, it's my third company.

" There is a way to simplify and monitor and measure how the world can be represented in terms of if I'm a customer and I'm trying to buy a product, or Toyota and I want to manufacture cars in the most highest quality way and efficient way, and how I need to leverage people and value stream management to simplify all that detailed level process engineering that's next level down. But how do we bring it up to a high level that's actionable for people? And I started applying that in my career, I want to say about 15 years ago, before I even knew that that's what it was, value streams.

But lean, systems thinking, the agile software revolution, fundamentally, they all start and end with the object of the value stream. I love it. Good stuff.

Tell us a little bit about Enverus. So Enverus is almost a product or a solution, I would say, by accident. So when I really started our first company, actually my second company, but my wife and I started a company called Techverus.

We started bringing agile practices, and this was in the very early days. We started bringing agile practices into Fortune 500 companies. Some of our biggest clients were when GE Capital was divested and formed its Synchrony Financial.

That was one of our biggest. I did some work at TD Ameritrade during a major integration and conversion. A lot of these organizations, we would go in and help them with their agility practices, applying lean, applying systems thinking, and it was through the process of doing that that I want to say about 10 years ago was introduced into value stream mapping and then value stream management, and a light bulb went off.

It was like all the stuff that we were trying to architect and build was really simplified through the flow metrics that we see today. And we started leveraging that for our customers and then building ways to cut down. As a services company, it might take a month to train, coordinate all the leadership you need, get time scheduled.

Do you spend a couple of days in workshops? And it became that architect engineer in me from when I started out early in my career was like, "Okay. Once again, we're in another left hand doesn't know what the right hand is doing and we can't get buy-in.

" And it became spreadsheet automation, and then it became uploading the spreadsheets into database And after a while, we started applying economics to it, a lot of the flow metrics that you see today. And then our customers were like, "Hey, can we keep that? " And I would say 2023, 2024, we started applying AI to the process as one of the first in the world to do that in a collaborative way, and we launched as a separate standalone company at the very beginning of 2025 last year, and we haven't looked back.

What a great story. Good for you. Yeah.

It's an exciting time. I know you asked, we're going to talk a little bit about Value Stream Consortium and the great work that Helen Beale and all of the real thought leadership that's in that community, and how we're really excited to be part of that. And what's happening with AI and this massive opportunity that humanity has right now to get out of its own way, if you really want to put it down into a term.

Because what's coming on the other side of the AI adoption curve right now is quite incredible. Yeah. Just before coming on this interview with you, every day at noon, we do the Techstrong gang for about 45 minutes or so.

Three segments. And with AI being what it is, they're always, in some way, shape, or form, connected. And just discussing today, we were talking about AI and data centers in orbit, how real or not that is, some of the deals going on.

The promise is so real. The promise is so big. The disruption capability is so great, how it can change our lives.

But at the same time, like everything else, that's a double-edged sword of not everyone will use it for good or in its best way. And it's why we can't have nice things on the internet. And it's going to be interesting.

It's certainly historical interesting times we're living in. Let's segue over, though. You mentioned the Value Stream Consortium, which is kind of giving way to the Flowtopia community.

And Helen breeds life into this every day. She manages it and everything else. This year's Flowtopia conference on the 24th is virtual, but really shaping up nicely.

I think at some level, the community does better not being a consortium per se, but more of a community. You are speaking there. Why don't you share, if you wouldn't mind, a little bit about what you're speaking of?

Yeah. And I think for this interview, it may have evolved. Everything is evolving really fast.

We have Moore's law for 50, 60 years as a proof of continuous disruption, doubling every 18 months to 24 months, and compute capacity on a chip. And with AI, we're roughly seeing an exponential or order-of-magnitude higher curve. And that curve right now is doubling about every six to seven months.

Today, you can download a public model, an open source model, and put it in your own compute, a Studio or a Mac Mini, if you can find them, and you basically have where the top reasoning models were just six months ago. And that's accelerating, and there's more players accelerating. So this discussion, I kind of called it the mind-the-AI gap, hearkening over and giving a little bit to our cousins across the pond in the UK and Helen Beale's work there with Flowtopia.

I still use Value Stream Consortium branding, sorry. But it really is the constraint. That gap is a value stream problem to solve.

The point of a value stream in its most excellent, purist first principle form is to align disparate functional groups that operate. It's selfish. I don't want to say that sinisterly, but it's human nature to bias towards your local concerns, and that doesn't improve the system.

And that's the gap that value streams can give us for, just like you said, Alan, if you fall on the positive side of the technology revolution curve, which I am, there's always that principle of creative destruction, which has lived in electricity. We inevitably put steamfitters out of business, but the transition was more job creation through electricity. It took us 50 years to adopt electricity from first commercial availability until the first factory was using electricity, and that was Henry Ford.

And we know what that did. Great. So all the way back to the Luddites and the automated looms, it's a continual story of humans advancing.

And so I tend to be on the abundance side. But yes, we're going to have to go through a lot of travails to figure out this whole AI adoption gap. Agreed.

So to me, this is where stuff gets interesting. The promise is there that a couple of things, how quickly it's going to come, how widespread adoption is. So does the adoption come in waves or what have you?

I think we need to really... It's going to be interesting is all I could say, right? Is how we do this.

And look, I'm going to be sitting in on your session for sure, because this is something I do find interesting. Besides your session, while I've got you on, any other session that you think our audience should be, like you as someone who's really in that value stream world, that you're looking forward to? Yeah.

So Steve Pereira, who- I know Steve ... he wrote Flow Engineering, has a really good talk on the human aspect of this, and I really love to lean into that. " Right?

So you're not selling the mechanics of doing value stream mapping. You're selling the value that you unlock. And so I think that human aspect of humans being on the loop and in the loop as we model value streams with AI is very, very critical.

So I'll probably be attending that one. And then a little bit biased, we have Devin Corona and Andrew Maurizio, who are engineering our AI orchestration loops to build our AI models, but also to take the workflow and apply that in an agentic way that orchestrates between mini models. Because you see all this stuff in the headlines about sticker shock on tokens.

Well, sure, when you're doing single loop interaction with a model and throwing all of your context at it, you're going to bankrupt your company. But when you're orchestrating across local models, mini models, and you're calibrating that to the scope of the work or the workflow, that's the critical piece. It's giving the inference model the ability to understand the workflow context.

And so I think we're going to see a recursive thing happening. We're going to see AI understanding value streams and using that to leverage the agentic work that we're trying to solve on this curve right now. So that would be another one.

They're giving a talk on AI engineering won't get rid of engineers, it's just going to make them more important. I love it. Yeah.

Hey, Brian, we're about out of time. I want to thank you for coming on Techstrong TV. More importantly, I want to thank you for presenting at Flowtopia, right?

Because the part about having a community is it's only as good as the community, and that means people stepping up, and thank you for stepping up. Well, Alan, thank you for the interview. It was really great, and I'll see you at Flowtopia.

Looking forward to it. Okay. io.

We're going to take a break here on Techstrong, gang. We'll be back in just a moment.