Subscribe
Share
Share
Embed
AI is changing how we work, but the real breakthroughs come when organizations rethink their entire foundation.
This is AI-First, where Box Chief Customer Officer Jon Herstein talks with the CIOs and tech leaders building smarter, faster, more adaptive organizations. These aren’t surface-level conversations, and AI-first isn’t just hype. This is where customer success meets IT leadership, and where experience, culture, and value converge.
If you’re leading digital strategy, IT, or transformation efforts, this show will help you take meaningful steps from AI-aware to AI-first.
Annie Baymiller (00:00):
One of my team members sits on each of the business presidents teams, so they're in the discussion as a business leader, not just an IT leader. And I think that's a real testament of technologies, not just for technology's sake, but it really does have to be tightly connected to how you want to enable growth for the company.
Jon Herstein (00:16):
This is the AI First Podcast, hosted by me, Jon Herstein, Chief Customer Officer at Box. Join me for real conversations with CIOs and tech leaders about re-imagining work with the power of content and intelligence and putting AI at the core of enterprise transformation. Welcome to the AI First Podcast, the show where we explore how AI is reshaping the future of content, enterprise workflows, and innovation. My guest today is Annie Baymiller, Senior Vice President and Chief Information Officer at Owens Corning. And I'm your host, Jon Herstein, Chief Customer Officer at Box. Annie leads IT at Owens Corning, a global building and construction materials leader. She's been spearheading digital transformation initiatives across the company, making her a particularly relevant voice on adopting AI in a traditional industrial environment. This conversation offers practical takeaways for CIOs and business leaders alike. So let's dive into today's conversation. Good morning, Annie.
(01:14):
How are you?
Annie Baymiller (01:15):
I'm great. How are you, Jen?
Jon Herstein (01:17):
I am doing really, really well. I would love to start with you just giving folks a little bit of a background on you, your role, and the scope specifically that you're responsible for.
Annie Baymiller (01:27):
Sure. Happy to. So I'm currently the CIO here at Owens Corning. I've really kind of grown up here at Owens Corning. I've spent the vast majority of my career here. I'm born and raised here in Toledo, Ohio, where Owens Corning is headquartered and joined in the early career program. Always have been in IT, largely on the project management side. So got to see how a lot of software was built and got to really grow from a kind of team leadership point of view. I left for a few years to do consulting and candidly really missed the culture of Owens Corning, which is really very special. We like to think of ourselves as low hierarchy, low ego, high collaboration, and I realize that isn't where the culture is for all companies. And so the stars kind of aligned as I was missing what I had left.
(02:15):
My predecessor in the role was building his leadership team as he became CIO. And I have been back now for 10 years and three years in the role as CIO. So I'm really accountable for all things IT, right? All things technology globally. So we have a very global team and we kind of talk of ourselves as kind of small but mighty. And we've grown significantly with a recent acquisition in the IT team has grown, which has been great. We've gotten some new perspectives, new capabilities, new talent who have joined us from Masonite, the doors company, which we purchased a few years ago. And yeah, collectively we lead all things from security to digital to AI to enterprise applications and recently the operational technology, so the technology and our facilities has rolled into my scope as well.
Jon Herstein (03:02):
It definitely seems like there's been a shift and probably you've seen it throughout your career from IT just being a utility, keep the lights on, keep the network running, that sort of thing to, as you said, really a key part of transformation and change and driving business value. Have you definitely seen that throughout here?
Annie Baymiller (03:19):
Absolutely. And I'm really proud of Owens Corning in that I don't think, at least in my tenure, at least from being more of a leadership role, there was never a fight to get a seat at the table. The seat was always open. And I think that's a testament to our executive committee of, while I think the clarity of how we add value is becoming more and more, but there was always good acknowledgement of we know technology is an enabler. We know technology's going to help us drive the strategy. We do a long range plan every year. And I remember years ago, each of the businesses did their long range plan and then I did the digital long range plan. And I remember thinking it's going to be really cool someday when there isn't a digital long range plan because it just sits inside all the business plans and we're there.
(04:01):
And one of my team members sits on each of the business president's teams. So they're in the discussion as a business leader, not just an IT leader. And I think that's a real testament of technologies, not just for technology's sake, but it really does have to be tightly connected to how you want to enable growth for the company.
Jon Herstein (04:17):
So where do you see the biggest opportunities for AI to start to transform how specifically industrial companies operate?
Annie Baymiller (04:23):
So for us to date, I've kind of just described it as it's almost been like a groundswell within the organization of where AI use cases have come from. So if I think back to the last couple of years, maybe 2023, early 2024 was really kind of testing the technology, understanding what's really going to evolve, figuring out how we build our infrastructure layer in a way that allows us to stay nimble as the technology continues to evolve. I think everyone's just kind of testing out, is this real? Is this going to continue to add value or is there, how much of this is hyped? And I think we all validated this is real. It's not going away. It's now time to unlock the power of it. And then if I think of maybe the last 18 months, it's been that, okay, so now we understand the technology and now we understand how the technology embeds into our stack in a variety of different ways.
(05:12):
What are the great ideas that we want to go after? And I think there was kind of just a natural curiosity in the organization where lots of people who love stuff like this, who love big technology changes, we're tinkering, we're learning. And in some cases it turned into a thousand use cases that I'll probably never know about where somebody took 15 minutes out of their day and that's great, right? Tell your peers, tell your friends, and then let's continue to be more efficient. And in other cases, it turned into these ideas that kind of built upon themselves where somebody said, "Wouldn't it be great if we could?" And other people got excited by that and we built some functional areas where we've really had kind of some meat on the use cases for sure. And now I think we're shifting into true value creation. Right now is the time for us to really go after how are we going to turn certain things into standard work so that we can either drive cost out, we can grow our revenue faster, or we can improve our employee engagement and retention by showing that we really are committed to creating a place to work where we use the best technology and get to the best decisions as fast as we can.
(06:15):
So I think there's this evolution of kind of lots of great ideas turning into now areas where we're going to send, we're actually going to change the way we work as a company and AI is going to be an enabler and that's the journey that we're on as we start 2026.
Jon Herstein (06:29):
I can definitely feel it. And we're seeing the same thing both internally at Box and then also with a lot of our other customers. So we'll go a lot deeper on the AI side, but I'm sort of curious for you personally, what drives you to lead change? And you think about both in a cultural and an organizational sense, what's the motivation for pushing for transformation at a company like Owens Corning?
Annie Baymiller (06:46):
Personally, I love change. I get bored when things are too static over time. But I also find that for me, the best way to lead change is to drop breadcrumbs along the way. And that's what we've been trying to do for the last 18, 24 months with AI is just keep it front of mind, keep it relevant, share little stories that are examples of someone winning and then feeling the benefit of it. Because as any large organization, there's pockets of people that love it and are excited. There's pockets of people that are afraid not just for what it's going to do to the industry, but just that they're not going to keep up and that it's going to be too hard and it's going to feel very different than how they're used to working. So taking the mystery out of it and taking the fear out of it by trying to just keep it as simple and practical as possible.
(07:34):
People don't need to know everything about AI and how it works inside the stack. They just need to know how it benefits them and how that they can change the way they work for the better. And so for me, that's what the fun of changes is watching little light bulbs go off for people and watching people who are maybe very uncomfortable become a bit more comfortable. And then I think you just get a natural momentum. And it's fun for us this year too, is our senior leaders from our CEO are very excited by the power of this. And so that energy and enthusiasm and really encouraging the organization to take a moment and think about how you work and think about where we could take waste out and where we could get to more efficient and better decision making, that level of, I think, senior leader engagement builds some comfort in the organization of we're going to figure this out together.
(08:22):
We don't have all of the answers, but we're going to continue to make sure that we're really maximizing the usage. Change is hard and it will be a multi-year journey. There's no light switch that's happening to enable AI for the company, but I think it will be a really powerful couple of years as we start to reimagine the business processes with AI.
Jon Herstein (08:40):
Yeah, totally agree. And you mentioned this idea of breadcrumbs. I'm sort of curious how many of the breadcrumbs are coming from that where individuals are saying, "Oh, I found this kind of interesting capability personally that I think I can now apply at work," versus how much of it is more top down you and other leaders driving initiatives and then leaving those breadthcoms for others
Annie Baymiller (09:02):
To follow? To date, definitely been the groundswell of ideas of people. Yeah, it's fascinating. Like I said, I love that there's a million things happening that I don't have to be pushing and I don't have to be trying to tell the story around because if all the great ideas are going to come from me, we're going to be short quite a few great ideas. But I think as you watch, one of the things we use from change management, which I have just loved is we do these sessions called AI and action sessions. And I jokingly say, think of it as kind of your seventh grade science fair to the max. And what we do is we do it both virtually and physically, but in the physical space here in our headquarters and some of our other big hubs, we have people with their poster board and they describe what their idea was, how they turned it into something with the help of either their peers or with my team from a technology point of view, and here's the value that is being created.
(09:54):
And sometimes the value is minutes and hours, sometimes the value is significantly bigger, and sometimes the value is even just around employee engagement and feeling empowered by the technology. And then we open the space up and people walk through and ask questions. And it's just a great way. For me, that's like breadcrumbs to the max is if you leave there feeling like, wow, I could go do something like that, or I should call that person and figure out how I enable what they're doing for themselves or their team makes it way less scary. And it starts to just show that we're taking pragmatic steps forward, but that ideation and thinking out to the future and being comfortable saying, "What if or could I? " It's really powerful, right? I think it builds this true groundswell of energy and enthusiasm. And because of that, I think it has encouraged and excited our senior leaders.
(10:45):
And now when I talk about that engagement, now it's worth kind of flipping that over now and saying, "Okay, from the top, we do want to say, we want you to reimagine how you work. We want you to think about business processes that we know have waste in them, that we know have too many manual touches, and let's think about how we reimagine them with AI and with automation and with more predictive analytics." And so that's the call to action going forward for the next couple of years is let's truly break some dishes and put it back together in a new way with some of these capabilities. And it's going to be fun. I'm excited.
Jon Herstein (11:17):
I love that idea of the AI in action sort of science fair setup. I'm just sort of picturing your deputy CIO running down to Kiko's at 2:00 AM to buy some trifold poster awards. You're not far off. So what are some of the coolest ideas you've seen coming out of that and have any of them started to make their way into true kind of production?
Annie Baymiller (11:38):
Yeah, a couple interesting ones. So we're obviously a manufacturing company, and so we're 85 plus years old with that. There's tons and tons of R&D data that has happened over the years, right? Lab notebooks, test results. And one of the questions that one of our kind of senior engineers in R&D had was, how do we leverage all of that data of all these years of testing and product development to make better, faster decisions going forward? Because there's things that continue to just the evolution, I think of innovation of things that will pop up every couple of years where someone will say something in related to how we develop a product that we've probably asked and answered
Jon Herstein (12:21):
Years
Annie Baymiller (12:21):
And years ago. And so how do we learn from what that outcome was and either acknowledge that it's the same outcome going forward or has something changed in the world or in technology or in our products that would require to be different? And so that's one of the areas that we're bringing generative AI is putting generative AI over all of those years of data from the lab notebooks and test results, digitizing ones that were physical to date so that we can interrogate all of that years of history and say, when was the last time that we evaluated using product X instead of product Y in our raw materials? And then seeing what all those results were and then being better going forward in terms of how we build. So ultimately wanting to build faster and at a lower cost. So that's one that we're really turning into production.
(13:08):
We're turning that into standard work that if you're in that role, this is a tool that is now something that we would consider part of our capability suite for R&D. Another one would be we're a company that truly believes in safety and we want every single person to come to work and leave in a better position or same position as when they came. And so we take safety incredibly seriously. And one of our safety leads said, when we have incidents that happen and someone is injured, are there things in the environment, whether it be time of year, what shift it is, whether we're sold out or not sold out, that are there correlations between those things historically when we've seen injuries so that as those things start to emerge, we can say, hold on, this is sometimes where we've had an issue and so let's be diligent in terms of how we lead through that.
(13:57):
So that was an example of, we were able to put AI on top of that, to look back at all the safety reports and all of the things that are happening in the macro environment and that specific plant environment and an opportunity then to be more predictive in terms of when we know we could be in coming towards an environment that could lead to an incident so that we can really lean heavily into the plant leadership there to avoid those. So that's one that plays out in real time as we look at the data of the past, which is really interesting.
Jon Herstein (14:25):
And that's an incredible one. And I'm curious if that sort of area of safety analytics is taking advantage of generative AI capabilities or is it more traditional kind of AI? And I guess what's new and different now that you didn't have before?
Annie Baymiller (14:39):
I would say the modeling is probably more traditional
(14:44):
Where we're looking at fundamental safety metrics that we can look at for more of a kind of computational mathematic point of view. I think what's different is the ability to bring in unstructured macro data and external data and then to be able to interrogate it in a more conversational way, right? As opposed to looking at numbers and charts, that's what we're continuing to build is now that we say, "Oh, this is interesting and this could lead to some really interesting outcomes." How do we continue to make it so easy to use that we get to those insights faster in a way that doesn't require you to be a math expert always to get to understand what the outcome really is. So that's why I think when I said before, I think it's this combining of the capabilities, the things that we've done and we're really good at and we trust the data with now how do I use the other pieces and versions of AI so that it just becomes so user friendly that we get to the outcome really quickly.
Jon Herstein (15:36):
What has your process been for thinking about digitizing assets that aren't digital today?
Annie Baymiller (15:41):
The biggest focus to date has really been on the R&D side as we've been thinking about all of that, just because I think the value in that space is limitless, right? I mean, there's so much we could do from how we build what we build, and then that ultimately leads into what the manufacturing process is. The linkage there is not just around obviously the innovation of the product itself, but the innovation then and how we build it. I mean, that's the core of the business is the manufacturing, right? If we don't produce the right products in the right way, the rest of the company is useless. I mean, that's our crown jewel is our manufacturing. So I think that's an area where we'll continue to see that linkage between R&D and innovation and then how we bring that into the facilities, both from a, as you said earlier, a robotics point of view, like where does that really add value, but then this right data in the right hands at the right times to make a decision on a plant floor, that's the secret sauce that we're continuing to work at.
Jon Herstein (16:37):
So when you identify an opportunity for a technology, and let's keep it around AI, since that's so top of mind, but whether it's to analyze data or automate a process, how do you decide, what's sort of the framework or rubric that you use to decide whether to build a solution internally or adopt a commercial product? We'll talk about Box in a second, but just generally, what's your framework for approaching that question?
Annie Baymiller (17:00):
The lens that we have been using is I want to be able to buy in the areas that we are not a competitive advantage or it's not a core to who we are as a company. So am I going to be able to build collaboration technologies faster than boxes building? There's no way, right? So it would be silly for me to go and try. And we're, as you said, a manufacturing company. So why would I go try and build something that fundamentally the thing that's out in the industry is always going to be better and you're always going to be able to invest more in those things than I could invest in that one singular product. So those are the ones where I would say, maybe I'm not going to get a competitive advantage from the technology itself. My competitive advantage will come from how do we use what we buy to the max and how do we make sure that we're really finding the use cases inside that subscription or purchase that allow us to win in the market.
(18:01):
The areas where I would build would be something that is so unique and fundamental to who we are and our data that I think I can build it in a way that allows us to unlock something that somebody else would not be able to unlock. So the capabilities around R&D and things like that, that's an area where our data is so unique to us and we know inside that data what really could add value different than I think anybody in just the tech industry would know because it's who we are as a company. Those are the areas where I think the build gets really interesting because we can tune it and we can train it and we can utilize it in a way that only we really understand that allows it to be a competitive advantage. So that's a significantly smaller scope of the build where there's the buy piece that ... I think there's a couple different pieces inside that too, where there's the areas like Box, there's areas like Copilot where there's going to be hundreds and hundreds of use cases that I'll never know.
(19:02):
And that's just a call to action for everyone in the company to feel empowered to go find ways to be better and find ways to work differently. And then there's the buys that are very specific to a function that would be a certain role where it'd be a bit more, I'd say, niche in terms of the outcome we're trying to create, but still one where we'd say, "Hey, if we have this tool and that provider is embedding AI into their tool and they're making sure that they're pacing, our job is just to utilize it to the max, I think, instead of try to compete versus building something unless it's a competitive advantage source for us."
Jon Herstein (19:37):
Have you seen the need for establishing net new roles that didn't exist before to kind of make sure that the AI is being fed with accurate data, up-to-date data, et cetera, or are we already doing that?
Annie Baymiller (19:47):
Nothing I would say net new for the company. I'm sure there will be over time as we continue to learn more. The way we've really approached it is two kind of different bodies. So we have what we just call our AI community of practice, right? And so it's really, think of it as a traditional hub and spoke model where the hub is my team from a team that can bring the right data, the right technology, and then the spokes are those curious, interested people in all the functions and businesses that are bringing the problem statements like, "Wouldn't it be great if we could?" So that's how that ground swell really started is these people that just were so interested in the outcomes, the technology, they knew they were going to get it wrong some of the time, they knew they were going to get it right some of the time, but they had that courage, I think, to go ask those questions.
(20:32):
That's how it really started with, "Ooh, okay, let's go test it out. Let's see if we can go do that. " And so that was the validation phase of the technology and the use cases. Now we're just growing that. Now we're just saying, okay, let's be really purposeful about whose name is in the box for each of the spokes so that that person is being rewarded in their goals and rewarded in their outcomes for driving those outcomes. So I would say it's just formalizing it in a way that we probably didn't have as much as we were in the learning phase. I'd say that's the ideation innovation funnel that now is going to just get bigger and bigger and bigger as we take, like I said, a more top-down approach of re-imagining the company. So everyone's going to feel like a spoke, but there still will be a structured hub and spoke model so that we can prioritize across the company.
(21:17):
And then the second group is what we call a responsible AI group, which is the group that takes those ideas before they would move too far into evaluation or build and make sure we're doing the right thing in the right way. So that's our IT security, our data architecture, our legal with representing each of the regions since data privacy and residency laws are different across the world. And so they're the ones that look at that and say, is this the right thing to do with the resources that we have and make sure that we're always got a human in the loop as we're building and even deploying. So those kind of two governing bodies and ideation bodies have been kind of the structural pieces that are in place. I do think in the next couple of years, could we imagine someone who wakes up every day thinking about data curation?
(22:03):
Probably. I think we've just got to continue to, as we really reimagine and go through this transformation with thinking about our business processes differently and how we would bring more AI in, I think it's going to inform us and over time of what we need.
Jon Herstein (22:17):
So I want to ask you one last question here, which is based on your experience so far, what advice would you give to other CIOs in traditional industries who are just beginning their exploration of AI? And are there mindset shifts, early investments, partnership approaches that you believe are critical in setting the stage for AI success or really anything that you think would be helpful for your peer?
Annie Baymiller (22:37):
Yeah. I'll give you three. One would be if you haven't already run fast and hard at your data, especially your structured data and get that in a position where it's accurate, it's trusted and it can be used. Because if that's not there, the capabilities on top of it are never going to be trusted. So now I think we all have work to do on the unstructured side and just even on the structured side, making sure we're using it the best way. But if you actually have some fundamental gaps in terms of your data architecture, I would run at that as fast as you can. And secondly, I would be purposeful and create those fundamental boundaries and guardrails so that you allow people a ton of freedom on some of the bigger pieces like a Copilot or like a big internal ChatGPT style. And so I find some companies who they'll leave some things maybe wider open and say, "Well, we're just dealing with it on policy." I think that's fine in some areas.
(23:40):
But for me, there were other areas that I said, "No, I'm going to lock it down now. I'm going to build something that's secure and still reaps the benefit of it, but requires us to do it more for OC because then I know there's no way somebody could make a misstep when they didn't mean to, but they were either putting data out there that we wouldn't want. " So set your own guardrails around the technology that allows the innovation to happen at scale for everyone. And then three, just storytell it all day long. Don't wait for perfection where something has finally reached the outcome two years from now around revenue or cost takeout or efficiency. Celebrate the little milestones, celebrate the ideas that turned into proof of concepts that maybe still will get killed because they don't add value. But it's the only way I think that it builds this sense of empowerment and accountability for everyone to be in it and know that no idea is a bad idea in early days.
(24:38):
And then the storytelling just helps to bring the followership along of what we're trying to create and the goodness in it and what we know and what we don't know. There's a lot we don't know and that's okay. It demystifies it. So I just find ways that you can constantly be storytelling.
Jon Herstein (24:53):
I love that third one in particular. I don't think we talk about it enough, just sort of talking about it as you go, which I think creates more comfort for people, particularly comfort in this idea that you referred to of like, they're not all going to be wildly successful and that's okay. This is moving too quickly to say everything's going to be perfect and everything's going to add value. It's sort of a Silicon Valley mantra of fail fast. And I think this is exactly where you are with that and telling our story.
Annie Baymiller (25:21):
Sometimes when too much time goes by and people don't hear the evolution of the space,
Jon Herstein (25:29):
I
Annie Baymiller (25:29):
Think there is a natural ability to say, "Well, something interesting must be happening and they'll bring it to me when it's done and then I will begin using it. " And I just think that's not how this is going to work. I think there's going to be some of those where we're going to say, "Yeah, we are going to build a new capability set for this role or this function," but I also think we need 25,000 great ideas every day of how the processes can be reimagined at the doer level, at the connection level that I'm never going to go identify. I need the people who live it every day to say, "This is wasteful. Wouldn't it be interesting if I could do it a different way?" Or, "Man, I wish I had this data at my fingertips that I could use in my decision-making process and then how would I automate my decision-making process?" And so if there's too much mystery and space between talking about certain things,
Jon Herstein (26:19):
I
Annie Baymiller (26:19):
Think it pushes people back to the bench waiting to be called on versus, "No, no, no, these are the million ideas that are happening every day across your peer group and it's the time to raise your hand and be part of the solution."
Jon Herstein (26:31):
Right. Participate versus waiting for it to show up. It just feels so different from the old days of the ERP replacement project that was two or three or four or five years in the making. And we'll let you know when we're rolling it out kind of thing.
Annie Baymiller (26:44):
Exactly. Yeah, this is very different.
Jon Herstein (26:46):
Very, very different. Well, you've done a great job explaining a lot of things to folks here, including some of the differences and how people should be thinking about it. Thank you for the advice. Thank you for your time.
Annie Baymiller (26:56):
Thanks for having me.
Jon Herstein (26:57):
Yeah. And Annie, also for many, many years of partnership with Owens Corning, we really, really appreciate it. And to all of you, I would say thank you for listening to the AI First podcast. Again, sincere appreciation for Annie for contributing to today's discussion. And folks, if you found value in this episode, please subscribe, share with your professional network and join us for our next episode as we continue exploring artificial intelligence, content, and enterprise transformation. Bye for now. Thanks for tuning into the AI First Podcast, where we go beyond the buzz and into the real conversations shaping the future of work. If today's discussion helped you rethink how your organization can lead with AI, be sure to subscribe and share this episode with fellow tech leaders. Until next time, keep challenging assumptions, stay curious, and lead boldly into the AI first era.