theCUBE + NYSE Wired: AI Agent Conference | Mikkel Holm, BARK

Clips
More from theCUBE + NYSE Wired: AI Agent Conference

Mikkel Holm

Chief AI & Innovation Officer

BARK

play_circle_outline Implementing AI across all business departments

play_circle_outline Focus on business transformation and team makeup

play_circle_outline NYSC Wired Community Event: AI Agent Conference Features Diverse Range of Speakers and Participants in May and June

Info
Transcript

Mikkel Holm, BARK

Mikkel Holm

Chief AI & Innovation Officer BARK

Mikkel Holm, the chief AI and innovation officer at BARK, joins John Furrier of theCUBE to explore the transformative journey of artificial intelligence at BARK during the AI Agent Conference 2025. In this session, Holm discusses how AI reshapes operational and creative processes across the organization.

Holm brings a unique blend of creativity and technical insight to their role, championing AI integration from product design to customer engagement at BARK. Hosted by Furrier of theCUBE, this discussion uncovers key topics such as the challenges of cha... Read more

explore Keep Exploring

What is the main focus of the individual's job in relation to implementing AI across the business and how do they approach it uniquely? add

What factors need to be considered when putting together a team for business transformation and dealing with ambiguity in the process? add

What can be expected at the upcoming AI Agent Conference in May? add

bolt Powered by CUBE AI

Mikkel Holm, BARK

search

>> Hello, I'm John Furrier here in our Palo Alto CUBE Studios. Hosts of theCUBE are here for a special presentation for the NYSE and CUBE special coverage of the AI Agent Conference happening in May. We're featuring some of the key speakers and innovators making the real world applications that are driving this next wave of innovation around AI experiences, and of course, looking at the ultimate outcome, which is what's going on in people's lives. Mikkel Holm here, Chief AI and Innovation Officer at BARK. Mikkel, thanks for coming on theCUBE. Great little preamble before we came on camera about what you're working on. Can't wait to get into it. Thanks for coming out. Explain what you do real quick. I want people to understand the context for the conversation.

Mikkel Holm

>> I'd say the core of my job is implementing AI across the business, and that means from every department, from design to finance, to supply chain and customer support.

>> You have a unique approach to AI. Obviously you got Chief of AI and Innovation, you have a creative background and what you guys do as a service is unique because you're using the data and you have the system set up. You're taking it from a creative perspective, but also looking from the practicality. Explain what you're working on, what's the environment like? Let's set the table.

Mikkel Holm

>> Yeah, yeah. I come from creative myself. I ran our product development and creative for many years and what we've seen moving into AI and trying to implement that into the business, we've seen it much less as a technical challenge and much more of a change management or human problem, if you will. And we've seen this and we've all heard the stories of it. Every time the conversation goes to AI, it always gets to this, "Oh, am I going to lose my job? Is AI going to take my job?" And I've changed my position on it where we used to have this phrase of saying, "No, you're not going to lose your job. You lose your job to someone using AI." And recently changed it a bit to, it's like, "Yes, of course you're going to lose your job. Of course you're going to lose your job to AI, have you not been paying attention at all?" But what this little caveat is, "Wt least the job as you know it today." So our approach to it is thinking across the organization every single role and asking the question, "How is this going to change?" It's not going to stay the same. It's not enough to just, in our case, we make silly plush toys for dogs. It's not enough just to add AI on top of a job description. It's like, "Oh, now you're AI enabled." It's much more interesting to say, "What can that toy designer do now that they have AI tools available?"

>> So as you're thinking about it from the perspective of, "Yeah, your current job is impacted, but the new job opportunity frees up for you." So you're soliciting input, is that what happens next? How does that go? Take me through, by the way, I love that approach. Take me through that process because again, change management is part of transformation, It's changing. What is that conversation? Are you soliciting input, "Hey, what do you want to do with free time or more cycles of brain power?"

Mikkel Holm

>> Yeah, I think we often get to that of like, "Oh, don't worry. Now you can do more strategic work." It's like not everyone needs to do strategic work. That's not a thing. So the way that we are talking about it in the way we're approaching is I have good allies in every department. So I have 10 people across the organization who are my allied. I've jokingly call them my spies, but the idea that they're in the team and they have deep domain expertise, they understand, let's say supply chain. My colleague Brian Grochal there, he understands deeply all the pain points, all the bottlenecks, all the workflows in there that are stringing that department together. But he's also deep in AI and understands the tools and what is coming and what we can do and what we are doing then together with Brian, let's say in supply chain, is running a bunch of small experiments like, "Hey, this little part of the workflow we can automate." And I think oftentimes it gets to this sort of grand promise like, "We're going to change everything and we're going to change it overnight." And it's like, "That's not how it's going to work." We're still running a business. We can't stop. Everything's moving at full speed ahead.

>> It's interesting because I think that's a practical approach, but the holistic change happens when you chip away at it. So take me through that innovation strategy. Is it you got some departmental focus where the domain experts are? Is it to identify bottleneck workflows? What's the approach? Is it process improvement meets creative? I mean, how do you view that and share your thoughts? Because this is what people are trying to figure out now, which is I don't want to scare people to death that the world is just going to drop the answer on the doorstep, and we now have AI and that's clearly hype and that's not going to come overnight. It's a lot of work.

Mikkel Holm

>> Yeah, it is for us. And I think just like everyone else, we're also trying to figure it out. And I think the more I talk to people in similar roles in other companies, they're like, everyone's trying to figure it out. Our approach has very much been this, as you say, a lot of experiments and trying to string those together. And because also the definition of experiment is you don't know what the result is. If you know what the result, it's not an experiment. So a lot of these are not going to work. We're trying a lot of things that won't work, but it's very core to us. That's our thesis for winning at the moment. Back when we started a company many years ago, we were early on growth marketing, digital marketing, and early on all the platforms and did a lot of experimentation there as well. So we're seeing this as another one of those moments where there's a moment in time where we can go up against bigger companies, stronger companies if we move fast and we experiment.

>> Talk about the impact on the key use cases you're focused on. Is it product development? You mentioned supply chain. Is it operational? Is it product? Is it customer? All the above? Take us through some of the things you got your hands in.

Mikkel Holm

>> Yeah, it's all the above. And I think that it falls into two categories. One is the internal processes we're talking a little bit about, we'll go deeper into that, but there's also external facing like, what are we doing towards the customer? But for the internal part, we are have these group of 10 people across the organization who are helping us identify. And we are at the moment going quite deep on supply chain. Some of those examples are very small, but just in organizing information, there's so much information going in and out of our supply chain team every day with partners and shipping, and especially in this time of the shipping market being complicated, a little bit extra with the tariffs coming in. So one of the things we've doing there is just having some of that data transformation and just take from unstructured data into a dashboard that we can act on a daily basis.

>> Is there a playbook or personnel or team approach you see with this transformation? Because AI, when we look at all the interviews with theCUBE and the research we're doing is that we've been talking digital transformation for over a decade. Rah, rah, check, check, check. Now the conversation is business transformation because a lot of business model things are involved, CFOs are involved. You're talking about domain expertise, you're talking about people close to the customer, close to the problems, close to the operations. You're seeing a lot more business operations. So the team makeup becomes a conversation. Some people like to deal with ambiguity, some don't. How do you look at the team makeup, the discipline required, proficiency? Can you share your thoughts on folks that are trying to put the teams together or have their departmental spies or access data? What's your approach on that? Share your thoughts.

Mikkel Holm

>> I think you're right. The question that we have and trying to figure out is how should we structure ourselves overall if we were starting over again? If we were starting the company today, how would we approach it? And unfortunately we have some of these silos built, if you will. That's a bad word, but some of these group built and defined already and they're running the day-to-day business. So the challenge is how do you insert yourself into that? Because I think a lot of it would require breaking some of these down. As example, the split between marketing and growth, marketing and creative. It has often been they're working together, but I think a lot of the smart marketer of the future will be able to create some of that creative themselves or the smart creative will be able to actually bring something to market, do a media plan, and actually analyze the data themselves. So a lot of those roles are shifting. What we are trying to aim for is there's a couple of layers. One is the basic is the automation. There are some of the jobs and some of the roles that are going to be automated, and it might be a little by little. In some cases it's the full job that's going to be automated. More and more are coming on that and it's getting easier and easier to build all these agents automations. The second layer is more interesting where it's in augmentation of our team. So I think the beginning when I was starting this journey, I think everyone at the company had this idea, I have list of everyone's name and I strike through and write AI next to them. That's not the case. What I do like with that list of all the names of the employees, what's that personal AI tech stack, if you will, that's next to them? What are the tools for each individual that are right for them to enhance how they work? And so we can augment every single team member. Because it's great If we have few of us that using the tool and getting, I get 20% more efficient. But if I can get 100 or 500 people to be 2% more efficient and it's starting to get an impact.

>> I love that phrase personal tech stack, because that's your toolkit. It's totally awesome. I haven't seen anyone really kind of standardized on a toolkit. You've got ChatGPT, you got Perplex, all these different tools. I see memes all around. Here's your toolkit, but that's where it's going.

Mikkel Holm

>> I think that's where it's going. And I think the other challenge is this is not a static, because it's going to evolve so quickly. Very specific recent example, our design team been working with different kind of image generation models, and then ChatGPT drops their new image generator and it's just so much better than what we've been doing before. So it switches workflow literally within the past few weeks. So I'd say yes, we need to define that toolbox and that tech stack, but it can't be static because it's still evolving so quickly.

>> Yeah. One of the things I've been riffing on with folks I've been chatting with is that what's the human intelligence changeover? You had a creative background all your life, you're in operations, was seeing almost an interdisciplinary convergence on the human role, human in the loop or whatever you want to call it, human in the loop, human intelligence, human creativity, human efficiency, collaboration. You starting to see a bunch of, I don't want to use the word soft skills in a literal way, but almost a soft skills, but interdisciplinary, kind of like a new world courseware coming that's required for people. Because if you're a static one tool player, maybe you might not survive. What's your reaction to that?

Mikkel Holm

>> I think you're right. I think I like the soft skills, because I do think the other word for the we've been using has been orchestration. So you're the conductor of this. So what you need to be there, that's the leadership role that is conducting all these tools and all these processes, whether you are as a leader, you have a mix, I think of humans that are working with you, but you also have a mix of in their agents or full workflow automations and knowing when to do what. So some of those skills I think goes back to classic leadership skills of understanding how to interact critical thinking, how to interact with people. Some of those are going to be static, but I think you need a very flexible mindset to how you get it done.

>> I noticed you have a background at Lego, and I want to bring up a question, not a question, a comment. I'd love to get your reaction. I've been seeing the engineering side of the trends where whether it's coding, AI infrastructure, I mean, certainly in one area, the engineering side, all the alpha coders are going down in the stack to the firmware level, microcode, assembler, they're squeezing every performance out of the chips, a whole other systems design thinking going on around there. And then as you move up the layer of the stack, but there's a certain part of the stack and below where the engineers are like Lego engineering. I mean, they're like, I've used the word Lego, the Lego engineer is the new engineer because there's a lot of pre-built things and coding assistance going on. So you've been there. So since I use the word, do you agree it's kind like a Lego block kind of vibe here? And then if anyone wants to build their own Millennial Falcon, they could do that. So again, the creativity comes back to full circle. What do you think about that analogy? Because when I was in school, you didn't know everything. I had the code up and down the stack. I get to go to the operating system level, do the graphics, but now with open source, you're starting to see an engineering mindset that looks a lot like the Lego generation.

Mikkel Holm

>> Yeah, I'm a big Lego kid still to this day as you can imagine. So I like any kind of reference like that. I do think you're right. Take myself as an example, I've always been an idea person and I've always been dependent on other people to bring it to life. I've always been dependent on someone, whether it's the engineering team or the media buying team or whatever it could be. And I find out what's super exciting is that now I can bring a lot of those ideas to life myself. And I see a lot of entrepreneurs doing this and much smaller teams where you can do a lot more yourself. As you're saying, the building blocks are there. I do think there's the other side of the mix to it. You still need domain expertise. You still need to know and the people who are really taking advantage of these building blocks are those who know what good looks like within that domain. So that's why it was internally for us, when we are working with the supply chain team, I can pretend to be a supply chain expert for a while with ChatGPT behind me, but at the end of the day, there's something of a domain expertise that's built up that you still got to lean on.

>> You bring up a great point. I love this riff conversation because creativity and craft becomes a huge asset. Having craft. You get some output from ChatGPT, I still got to package it. So there's a little bit of craft coming back, craftsmanship coming back to everything. What's your reaction to that? What's your view on having craft and make it human?

Mikkel Holm

>> I think it's right. It's the conversation that we've before of if there's going to be more and more AI-generated content, what's then going to be valuable? And there's this spectrum where, as we're talking about these days, you got to be at either end of this from a creative, again, point of view. Either you are fully automated and scalable and you're just generating fast content and output faster than anyone else. Or you've got to be truly original. And it's not saying original can't be generated with the use of AI tools, but there's something that's going to set you apart. And there's that big middle section of mediocre fine work that's just going to die.

>> .

Mikkel Holm

>> And that can be automated. No one's going to... So that's where there's going to be definitely going to be a displacement of jobs or roles at least. And you got to figure out how do we get to that truly original idea? And back to your idea, that's the craft, that's the ingenuity. That's the sort of a non-obvious ideas that we're chasing because it's just fun or just exciting that isn't in any training data anywhere because it's never been done before.

>> Yeah, that's the imagineering. Mikkel, thanks for sharing that commentary. I want to ask you, what's the coolest thing you're working on right now? Share with us the coolest thing you're excited about, coolest thing you did or doing. What's cool on your world?

Mikkel Holm

>> I mean, I'm such a nerd. I get excited about very simple things that are small automation. So there's a lot of small internal tools that we just excited about having. But one of the things we're really trying to do is more on that customer-facing side of it where we can deliver a better service to our customers or giving them new experiences that never had before. So there are obvious ideas of turning your dog into a dog in a certain style, but what we really want to do is bring your dog into the product experience. Every month we do these themed boxes. Next month, it's Tokyo. It's a spoiler. Tokyo is the theme of next month, and we want to bring your dog to Tokyo. So if you give us an image, we are going to insert it into our design and our creativity in there.

>> Hey, if you can bring a digital twin to that world, you'll have put Pokemon to shame on all these kind of virtual worlds. I think that's a great creative direction. I mean, that's an experience that's first generation.

Mikkel Holm

>> Yeah. The more we can do and the more we see that we get engagement from our customers, it's also a little bit of, we get more information with each step of this and we can service them better. So if I ask you for a photo, I get a little bit more information about your dog. If I ask you for the name, their age, all of these details, and the more and more I get, the more I can deliver a better service back to you. So ultimately it goes back to that first mantra, it's all about making dogs happy.

>> Owners are dog obsessed, as you know, you mentioned that. So Mikkel, thanks for coming on theCUBE, Appreciate it.

Mikkel Holm

>> Thank you.

>> Looking forward to seeing you in New York. I'll be there in June. We'll see a lot of CUBE action at the NYSC and of course the NYSC Wired community. A quick 30 seconds or so left. What's going on with this event? Put a plug in for Simon's event. A lot of thinkers coming together, a lot of real talent, real technical community, good cross-section of innovators and thinkers. The AI Agent Conference in May. What's it all about? What's going on?

Mikkel Holm

>> It is an amazing collection of people, and I am humbled by being to be part of it. I'm very much looking forward to it because as I've seen the lineup of speakers on participants there, it truly ranges from people like myself who are implementing it day to day into businesses, to people who are building platforms, to deep technical challenges being solved. So very much looking forward to learning a lot more.

>> I love how we curated that event. Again, AI is craft curation, imagineering, talent. Mikkel, thank you for coming on and sharing, Chief AI Officer at BARK. Appreciate it.

Mikkel Holm

>> Thank you for having me.

>> Okay. I'm John Furrier here for the NYSC Wired community and theCUBE, CUBE plus NYSC having a special presentation digital event around the AI Agent Conference. Thanks for watching.