Data Management Challenges with Panzura’s Dan Waldschmidt

This is Textron tv. Hey guys, thanks for the throw. We're here with Dan Walsh Schmidt, who's the newly appointed CEO for Panzura, and we're talking about storage and data management, and how this is all gonna evolve and this year and hopefully the next.

And as we all know, it's been an ongoing struggle for a long time, and hopefully maybe things are about to get better. Dan, welcome to the show. Hey, thanks for having me, Mike.

So what brought you to panzura? I mean, what left out at you here in terms of this opportunity and what, what's your assessment in the current state of data management? So, Panzer was attractive to me because of the core technology behind it.

This goes way back to 2019, when previous leadership team, um, knew that they had, um, some great IP but weren't exactly sure how to deliver it at scale. You know, a lot of hybrid cloud, multi-cloud technology has only evolved or, or really come into its own in the last 18 to 24 months. We certainly haven't crossed the chasm on, on hybrid multi-cloud technology for the enterprise space.

We haven't, there are a series of tools for SaaS, uh, SaaS tools for MEDLAR and SMB, but when I saw the technology for the enterprise of hybrid cloud that Panzura had it, I immediately got hooked on it. Um, and so it, it came as a, and as a consultant back in 2019, Mike. And then when the company went to raise money in 2020, uh, Jill and I, uh, uh, decided that this would be just the best use of our time to dig in and make this a company that we were part of acquiring and then scaling.

And of course, with the recent change in the last week, last two weeks, I guess now, um, you know, it's, it's now us taking this company in a, in a direction where stability and performance for enterprise customers, uh, is our number one focus. So, uh, what, what I fell in love with in 19, and what I'm, what I'm still excited about now is this unique ability we have to move data in and out of the cloud with almost an unlimited number of workflows and workloads, uh, more securely and a little bit faster than anybody else in the world. I feel like we've been talking about hybrid cloud now for a decade, and when I go talk to folks, they're still at the point where 90% of their workloads are running on one cloud and 10% are running somewhere else.

And so they'll check a, uh, multi-cloud box, but it's not really hybrid cloud in the sense that things are not centrally managed or unified. So what are we gonna do to get there and how does that conversation start with the storage systems? You know, when I, when I'm at my house, I see my kids watching these cooking shows.

Uh, you know, when I'm not on the road visiting customers or meeting our engineering teams, I see 'em watching this and this one show like the Great British Cooking Show where they make, you know, cakes and pies and all kinds of neat things. When I watch those shows, I, I get frustrated because like, I, I see them making something, I just don't know how they put it all together. Like, is there a, behind the scenes, is there a recipe book that those, uh, participants are using?

You know, I, I feel like for hybrid cloud, we need a recipe book. We, we need a bit of a cooking class on what's possible. You are seeing multi-cloud adopted faster now than hybrid cloud because of some of the unknowns around hybrid cloud.

What's been exciting for us in the enterprise space is how we're beginning to build this cookbook. I think we might be at the tip of the spear with workloads like HPC, where previously you were either cloud native and had to do it all in the cloud, or it had to be all OnPrem. There's some neat recipes that are being built and used at an enterprise scale around making HPC truly actionable in hybrid and hybrid multi-cloud environments.

So I think there's a lot of theory around what can be done in hybrid, but when you're asking CIOs to spend their money on investing into hybrid, I think that's where you're seeing a little bit of reluctance or retraction from that. But as, as leaders see workloads being able to be more stable and more performant and have access to cloud services like SageMaker and AI and all the different, uh, capabilities that are now available in multiple clouds, they're now leaning in to, to start building this. But I suspect Mike, like, like you've suggested, we're probably another 24, 36 months away from this becoming just an everyday enterprise practice of using hybrid cloud, uh, the way that we envisage it could be used.

Do you think during uncertain economic times, more people are concerned about getting locked into a particular cloud provider? So now they're more willing to look at what's required on an architectural level to give them flexibility More than ever. More than ever.

Right. It's, it's it, and it's not just the cost. When I meet with CIOs, they're also trying to discern which of the services are gonna truly be most valuable to them.

For example, ai, it felt like AWS was at the front of the pack with ai, but with open AI and the relationship with Azure and Microsoft with open ai, all of a sudden there's no questions with your enterprise data, where are you going to, you know, consider deploying AI services? So it's most definitely the budgetary insights are make a difference. It's also who is going to take that next step forward.

We see this in a lot of industries. You know, automotive, someone puts an autopilot in their car and then the next, uh, you know, manufacturer has to do it. We, you know, we see those additional capabilities when in lots of different industries cloud's.

Interesting because unlike a car where you could sell it this weekend to get a new one, once your data's in a cloud, there's not many services that allow you to move it from one cloud to another cloud without breaking the bank on costs. What is the relationship between on-premise and the cloud these days? 'cause when we talk about hybrid, I would include those environments as well.

A lot of organizations still have most of their data sitting in some data center that they manage somewhere. Can we federate the management of data across these environments? We have to.

We have to. That's where, that's where we're seeing the industry go is you CIOs are reluctant to invest in hybrid when they just don't have a clear status on where is my data and what it's doing. Um, Mike, I actually think this is a great use of ai, um, for the enterprise.

There's a lot of AI being used, um, creating shiny, uh, sort of widgets for medlar and SAB. Can you summarize a file? You know, can you extract information chat, GPT format from a file?

Things like that. But awareness and interpretation using, uh, AI to, to bring back information to a CIO on what is in those data silos, I think is a, is an, is the next obvious step for enterprise companies like us, uh, to be able to give information to a CIO that lets them make those decisions around, are we gonna keep this on prem? Is this a great workload for hybrid?

Um, could this be hybrid multi-cloud allows 'em to make those decisions smartly As we kind of embrace ai. At the end of the day, it's a lot of data that we're gonna use to train these models. Do you think that the rise of AI is gonna force a conversation around data management that previously we kind of, you know, ignored because, well, we weren't dealing with this level of scale, but now it's unavoidable.

If I'm gonna train something I need, well, what at least petabytes of data eventually to deploy something that may only be the size of a couple of terabytes, but at the beginning it's huge. The rise of AI is forcing us to rethink storage and data management, right? At a bigger level, what's behind ai, which even has geometric impact on this, is quantum computing.

Because if I can run a thousand parallel processes in a quantum computing model, what does that say for my data? There's a lot that we're still digging into around the implications of both AI and quantum computing on the future of data management. I think about this through the lens, mike, of data enablement.

So if I can store it, that maybe is the most basic of operations at an enterprise level, if I can manage it, that's a little bit better. It's providing a higher quality of care to CIOs. But what if I could truly enable it?

One of the visions we have at panzura is being able to empower employees to do their best work, serve up the data they need, where they need it, even while they're thinking about needing it, right? Serve up CIO's answers before they think they need it. Empowering the organization with tools and productivity that previously required manual processes or tools or, or automations that someone had to think about.

How can we be one step ahead in delivering that at enterprise scale? Do you think in a lot of ways, as we kind of quote unquote all discover AI, that we're gonna rediscover that a lot of the best practices that were defined for high performance computing, A-K-A-H-P-C apply, and we kind of already know how to do this. We just, it's just been in the realm of a smaller subset of people and now it's gonna go mainstream.

HPC, the, the, the, the regulations, the, the lessons learned for it around HPC are definitely line up to what we're seeing with ai. What we're seeing is that, uh, while Medlar customers and SMB customers are, are quickly trying to use tools like open ai, CISOs and CIOs, chief data officers, chief digital officers at large enterprise companies are, are, are rooted in the same fundamental, uh, principles that guide HPC, which is privacy and security. And then of course that has to be wrapped in a, in a layer of stability and performance.

But just putting your data into even an on-prem AI model leaves a lot of questions, uh, to, to, to be asked for security professionals inside the enterprise. What does this mean if my competition gets access to this data? If I can't regulate exfiltration and I've got a comprehensive AI model on-prem, what does that mean if I have an angry employee who might move some of that, uh, strategic intelligence outside the four corners of my office?

So you're, you're, you're spot on, Mike, in the, the sense that the lessons learned for HBC are, are definitely the bedrock or how we're building ai, uh, security and privacy models as well. Who's in charge of all of this? Because historically you had storage administrators running around, somebody else probably managed the data, and then we had cloud architects and DevOps teams we're provisioning cloud infrastructure resources.

Um, do we need to take a look at how all these IT teams are structured and maybe rethink it Well, uh, in, in the enterprise? And I just came back from a, uh, a week of, of customer meetings. I'll tell you, everybody has a say when it comes to, uh, to enterprise data enablement.

But maybe that's for the best. Maybe that's for the best. Everyone needs a little leads their slice of the pie.

It's a confidence game and not in a, um, you know, not in a negative sense of the word. CISOs need to feel confident that security and privacy is being handled. CIOs need to know that they have extensibility and that they're building on a platform that allows them to extend current enterprise capabilities into the next decade.

So maybe it's good that we do bring everyone to the table and step out of the shadows of, of everyone doing their own IT systems and everyone having their own separate data storage silos. It, it's a really positive thing to see the sort of job titles molding and melding right together. Uh, you know, your CISO role, your, your CIO, your chief digital officer, chief data officer, a lot of this, you know, it, it, the first question for me often in shaking someone's hand is Tell me how you define your role, right?

As chief data officer, chief digital officer, where the boundaries, what, what are you now responsible for? And then that informs sort of the collaboration that happens moving forward. It's a, it's a, it's a really positive thing, Mike, to see more people caring, not just about what's the bottom line cost, but what if all of this information I'm working on as a company, what if it's not just the, the projects and the decks and the financial numbers, but what if the, the operation itself, like the engagement of my employees allows me to scale as an operation, right?

That's my strategic advantage is having everyone at the table and then turning all of that collaboration into the next generation of who we are as enterprise leaders. So ultimately, what's your best advice for organizations to kind of get this ball rolling, short of maybe throwing everybody in a room and locking the door and hoping that reason prevails? Is there some way to go after this that kind nudges people in the right direction?

Yeah. Uh, you, you bring up a really good point, Micah, around what, how do you bring people together to make this decision? And I think the, the, the answer of the, is the, is the question itself?

What are we trying to achieve? I think that's important at the enterprise level. What matters to us most?

If you're an accounting firm or, um, a healthcare organization, uh, legal, you might value things like, you know, mission, cri, mission critical access to data might be more important than any new shiny whizzbang object you could build. But if you're an innovator and you're an enterprise, then you might care more about high risk, uh, data models than you would how securely we're storing our data. You know, the conversation itself, just the awareness that there's potentiality that we haven't tapped into yet is the starting point for this looking a se, the natural data storage cycles, the three year cycle, the five year cycle, that, that, you know, storage vendors over the years have sort of forced enterprise companies into, opens an opportunity for dialogue that we haven't had previously, right?

And allows us to say we're going to spend money. We either spend money smartly or we just go back to dump storage. And I think that conversation itself opens the door for us to have a dialogue around what do we value as an organization, and then what are the steps we can take to, to transform ourselves with these decisions that are really important to us.

All right, folks, while you're heard it here, every meaningful IT conversation starts with data and ends with data. The stuff in between is the means to the end. Hey Dan, thanks for being on the show.

Thanks, Mike. Back to you guys in the studio.