AI Is Accelerating How Fast We Build the Wrong Thing

Hey everyone, welcome back here to techStrong TV. We've been covering the upcoming Floatopia virtual event on June 24th. io.

But as part of that coverage, had the good fortune to interview and meet a lot of the people presenting. What a lineup of great presentations and sessions at this event. It's really a stellar event.

But when I saw this man's name, this gentleman's name on the list, I had a smile. We were talking off camera. Jeff Keyes.

com, and that was- Mm-hmm ... I don't know, a long time ago. 12 years ago, maybe more.

14 years ago. 12 years ago. And- It's been a minute.

Mm-hmm ... it's been a minute. But through it all, he still remained Jeff.

Hair was a little fuller then, but so was mine. Yeah. Other things have gotten fuller, too, which aren't good.

Exactly. But that's below the camera. Well, I've lost weight, at least.

Yeah. Yeah. But anyway, it's a pleasure to welcome my friend Jeff back to techStrong TV.

It's been too long, my friend. How you been? Life's great.

Family, boating, sun. Just watching the world of technology churn and all the new opportunities that are created. What a fantastic time to be involved in technology.

Yeah, it's historic. Jeff, I got to apologize because I haven't had you on here in a long enough time. I said Keyes, but is it Keyes?

It's Keyes. Yes. That's what I remembered.

And I said Keyes because I was looking at the teleprompter. I apologize. It is Keyes.

That's what I thought. All good. So yeah, it is historic times, Jeff.

It's interesting. Everything is almost being a little reinvented, if you will, right before our eyes. Mm-hmm.

And I think people being humans and humans having human nature, there's two kind of reactions, fight or flight kind of thing, right? Some people embrace the change, some people shun the change, resist the change. We've seen this through DevOps, we've seen it through Value Stream, we've seen it through Agile.

We saw it through Cloud Native and Cloud, and we're seeing it now with AI. Yep. It's a crazy thing.

But I wanted to just kick off a little, Jeff, not everyone out here has been friends with you for all these years, and they don't know who you are, so give them a little Jeff Keyes background. You bet. Look, I started off thinking I was going to be an electrical engineer who wrote code.

Instead, I ended up being a developer for many years through lots of startups. That turned into being involved as a product guy because I didn't want to be told what to do, which then involved I better improve my storytelling and the business cases of what we're doing, which led me through product marketing. And now I sit somewhere in the intersection between product marketing, product management, and what actually goes into working with customers from a customer feedback point of view.

The point is we have to get really good at what we deliver. AI, as it's changing all the roles, is making it so easy to build not the right thing. In fact, that was one of my favorite quotes, is from a CEO I talked to two weeks ago.

" I agree. Because the problem when you're on the rocket ship is if you're just off one degree at the speed, velocity kills, and it makes a big miss over short amount of time. Jeff, you're with Allstacks now.

Mm-hmm. I don't know if we've covered Allstacks a lot here at techStrong, and I apologize. But for our audience who maybe are not familiar, what's Allstacks about?

Allstacks has its foundations in what I would call software engineering intelligence. It's gathering metrics across the software delivery life cycle. Think of it like an observability solution for software development itself.

How productive are the engineers? What's the quality rate? How does it go through our overall throughput of delivering software and what it all implies?

Got its start in 2018 with a couple of guys that had their background in computer vision and used that, applying those techniques to create a really sophisticated context graph of all this data that aggregates from ideation to delivery. And when you mix that with code, it creates some really interesting things. Recently, Allstacks launched a new product that maybe we'll spend a little more time on while we're chatting here, but it's not what I'm going to talk about during the talk per se.

But there's a new module that we just announced a couple of weeks ago that targets the product manager to help them build the right thing and take off a lot of the weight of the minutia of creating the requirement and charting that out into tickets and creating the build-ready specs. It helps them really with the core problems of product management. So, that's the fundamentals of what Allstacks does.

We also have a module that does software capitalization for the finance side, for the tax benefit side. I love it. The website for Allstacks?

Jeff, the website for Allstacks. ai. Easy enough.

Yeah. Will work. Cool.

All right. Hey, Jeff, let's pivot. I opened this up talking about the Flowtopia conference.

Look, this is something you actually were instrumental in helping Helen Beal launch years ago now with the Value Stream Consortium and all of this stuff. You were one of the key players there. You've been involved pretty much ever since.

Mm-hmm. And thank you for that. You're presenting this year on an interesting thing, product management's new operating model.

Mm-hmm. I think people think, well, how is AI affecting product management? And that doesn't change, but talk to us about what you're going to talk about, Jeff.

Give us a peek. You bet. AI is having a dramatic impact on development, obviously, but for the most part, teams aren't seeing the greater throughput of what's being delivered.

What's happening as coding time decreases, there's other bottlenecks that are being hit. Theory of constraints, the bottleneck just moved. One of those primary areas is in definition, and what the product manager has to do is a quite varied and broad set of things.

The operating model for what new AI PMs will need to do is different than what they've been doing in the past. If you were a product manager that largely aggregated feedback from customers and put it into a big, long roadmap, some of that's interesting, but that's not really the primary job. Or if you were a product manager that looked at issues that were support cases to catalog those or write knowledge base or...

Those are some of the things, but that's not the primary purpose. The primary purpose of product management is to have a product taste and build the right things, and do it in a way that is sustainable, repeatable, can be extensible in how it goes, and help optimize the overall throughput to create the most amount of value that you can. One of the most interesting things that's happening is everyone thinks they're a product manager now.

" It looks great. Is it the right thing? That's the right question.

Is it the right thing? Does it have the right details? Does it have the right things go?

Designers are becoming part product managers. Hey- Yeah ... they can design this up.

Everyone's roles are starting to blur. The key problems that product managers face that are pushing them towards this new world, though, are a set of really a deficit in overall context, meaning aggregating all this data across all these different sources. And then second, a deep understanding of what's in the code to know what and how to build and how to drive engineering so that you get the most value out of your efforts.

That's the fundamental. What will I be talking about then under that context? There's a new operating model.

That operating model implies a different style of team, a different style of product manager. Does value stream management still matter? Kind of more than ever.

Because you know what? AI will go and build all the things again, even if you already have them, or not use common components, or build yourself into an architectural corner. Or you have to evaluate this using those same value stream principles to prove that the money and time that you're spending is in the right direction, is the most efficient, takes into account the kind of value that you want to look at, and ensure that you're doing the right thing.

The metrics still really help you from value stream management and will help you redefine how product management works within the confines of your team. I love it. It's interesting.

Some people are sitting out here saying, "But Jeff's the field CTO. " Of course, I know, Jeff, that you've run product multiple times. I have run product in a number of places, yeah.

Driven it for many years. Watched the impact of doing it right and doing it wrong- Mm-hmm ... in places, yeah, all across.

It really- And you have the scars to prove it. I do. Some of the favorite organizations that I've talked to have realized that their AI transformation shouldn't start with engineering.

It should start with product management. If you get the product management side right, product management will drive engineering in a way that everybody's aligned in the right ways to deliver value faster to their customers. Starts with that new shape of the organization, with breaking down these teams, with having fewer handoffs and more collaboration, with shorter cycles.

All these principles that we had called agile before are now really being realized in a way that I don't even know if they're agile anymore. It's sort of the new wicked fast waterfall that we bring together in this new world. And the value of having the fingers on the right set of tools to bring it all together, this is the killer use case for what AI can do for you.

Think of it as, if you're a person looking to understand what's going on, you've never had better tools to exhaustively analyze all the feedback, all the meetings, all the tickets, all the issues, incidences, so that you can create and design features that will have the highest possibility of success. And correlate that with the code, so that when it gets built, it'll work. Fair enough.

Jeff, we only got a couple of minutes. We've been doing so many of these interviews, but it sounds like a fantastic session. And really, look, you could talk about mashed potatoes and make it interesting, Jeff, but this is really something in your sweet spot, wheelbarrow, and I encourage everyone to go watch Jeff's session.

io. Sign up for this event. It's a great event.

" It does. If you're involved in running, in operations, in business, the model and everything else here, it does. Go check it out.

io. You don't have to sign up for the event right away. Go check out some of the content on there.

It's a circle community. They got a ton of great stuff, and you'll see, is this something that is interesting and important to you? And I think it will be.

Jeff, it's great seeing you, my friend. Likewise. Once Flowtopia's over, I'm actually going to be up next week at Platform Con, presenting in New York City.

But after that, let's make sure we get together. It's been way too long. Let's bring you back on.

We'll dig in more to what's going on with Allstacks and in your life. But until then, thanks for coming on here, man. Love it.

Thanks, Alan. All right. Jeff Keyes, Field CTO, Allstacks, presenting at Flowtopia next week, June 24th.

You're watching Techstrong TV