Conquering the Cloud: Panzura's Blueprint for Secure Data Management and Global Collaboration

Prepare to elevate your understanding of the cloud as the CEO of Panzura joins us to unravel the complexities of cloud security and data management. We're here to arm you with the strategies you need to ensure your company's transition to cloud storage is not only smooth but cost-effective and secure. Dan brings to the table a wealth of knowledge on the topic, offering exclusive insights into ways enterprises can tame the cloud beast, from automated data tiering to AI-driven optimizations. We're talking about serious savings and robust security measures that can make a difference to your bottom line and data integrity.

As we move forward, we turn our attention to the evolving battleground of cybersecurity, where preparation is the strongest weapon against the rising tide of ransomware attacks. Immutability and rapid recovery become our focus, as Dan emphasizes the need for systems that stand resilient in the face of threats. Our conversation with Dan also shines a light on AI's role in defining 'good behavior' within data systems and how it can revolutionize the way we protect and manage our digital environments. Furthermore, we explore how technology is breaking down geographical barriers, empowering data and collaboration to flourish in a post-COVID-19 global workforce. Join us to gain the critical insights that will help your business stay ahead in the ever-shifting landscape of cloud computing and data security.

Transcript

Speaker 1 0:02

Hey everyone, super exciting guest today and conversation around the future of cloud security, cloud data management and more with the CEO of Panzora, dan. How are you Doing well, evan? Well, thanks for being here. You're in beautiful Greenville, south Carolina, beautiful part of the country. Before we dive into our topic today, maybe introduce yourself and, for those folks who haven't heard of Panzora, tell us about your mission.

Speaker 2 0:32

Yeah, thanks, evan. Of course, Dan Walter, as you mentioned here the CEO of Panzora we're the undisputed first choice for hybrid multi-cloud data management. I don't know if you could get a bigger mouthful than that. Evan but that's the ground we've staked out and we're an exciting place where we're helping organizations, especially large ones, that are managing that move in and out of the cloud, needing the fastest and most secure path to move their data. That's the space that we own.

Speaker 1 1:14

Well, such an exciting topic, such a timely topic, as really everyone is moving to the cloud and storage in the cloud for so many obvious reasons, we understand the upside. What are some of the challenges, that pitfalls that companies, ctos, cios need to be aware of as they make this pretty fundamental transition?

Speaker 2 1:37

You mentioned to me when we were chatting or over commenting before we scheduled, this was cost. Cost is a big one. We can certainly come back to that, but like any new emerging capability whether now it's AI and pretty soon it'll be quantum the cloud itself has costs. It's not simple. The costs are not simple. They're actually. You most have to have a dedicated cloud mathematician to anticipate what your cloud costs are going to be. Cost is one.

Speaker 2 2:16

Security is a big one. If you manage to pay for all the right services, then you still have the possibility to have your data interdicted, intercepted, copied, shared, held ransom all kinds of bad actors trying to do things with your data. Security is a big one, then. Third, I think, is just getting the result that you want. We know that a significant number of workloads are being repatriated from the cloud back down to on-prem simply because whoever was making the decision to move the data might not have considered speed or security or user functionality, and now, all of a sudden, what you thought was going to be a great breakthrough for your organization actually turns into a whole bunch of open tickets that are hitting the support department.

Speaker 1 3:07

That is no fun. Been there, done that. Talk about some effective strategies that you've seen with your customers to manage those cloud storage costs, particularly at the enterprise level. Give us some tips, and tricks and insights. You must have seen at all.

Speaker 2 3:27

We're seeing a lot of it. I don't know that I've seen it all but, certainly a few lessons learned.

Speaker 2 3:33

There's four or five strategies where technology can now help you manage some of these things. You don't have to do it on a spreadsheet or a back of an app. One is automated data tiering and lifecycle management. There's more DLM capabilities available from the Cloud or just to stand on software providers. If you're not using machine learning algorithms to automate the process of data tiering, then your costs are going to be inflated. This is about dynamically moving data between different storage tiers in clouds and across clouds to reduce your costs basis on usage patterns. So one is automated data tiering. This falls under the data lifecycle management.

Speaker 2 4:21

Another thing that you would consider and should consider is serverless computing integration. Can you have a cloud utilization model that does not require physical servers? You can use hybrid cloud storage and you only play, only paying for the compute time. Of course, the genius behind the cloud is now. You're integrated into a service that doesn't need a server, and so, of course, you're sort of vendor lock in into the cloud. It's good for both the organization and it's great for the cloud. So think about how you can have a serverless computing model. How about AI driven storage optimizations?

Speaker 2 5:04

The third sort of trend we're seeing right now. You can leverage the current AI capabilities that are out there to optimize storage, allocation and performance. These are capabilities that you can consider before you're even moving that first workload into the cloud Is using AI to help you make decisions on where you should place data. And I think number four would be somewhere around the line of somewhere along the lines of multi cloud management platforms. So we used to talk about hybrid cloud, evan, and every discussion was what should be on prem, what should be in the cloud, and it feels like a lot of our conversations now are around okay, we're going to go to the cloud for most of these things. Which cloud? And so the conversation is really hybrid. Hybrid, multi cloud is really where it's at Now.

Speaker 2 6:01

This is just a unified view and control over different platforms and allows you to allocate workloads to specific clouds that have the best capability or return on investment for the spend. The best example of this might be just the current AI landscape, you know, being funded 10, 11 billion dollars worth of innovation with open AI. Of course, they seem to have a head start with their Office 365 integration, with co-pilot, and so you might choose to put your AI workloads right alongside your Word documents inside Azure, and you might decide to have some really snazzy you know database management inside Oracle's cloud, and then you might use Amazon for everything else, because they're everywhere and they do all things, and so you might want all three and you want them to work together seamlessly, and you don't want your databases on Amazon and you don't want your AI and Oracle, but you want all the data to be talking to each other, and so if you can manage that right, if you can be thinking about that right, you stand to gain from your utilization of cloud.

Speaker 1 7:16

Wow, my head is spinning. That is pretty fantastic and I think you answered my next question. But you know the hyper noisy, busy cloud, multi-cloud landscape. You know so many options. Customers, partners are probably confused about choices. How do you stand out in this noisy landscape? Again, the tech innovation clearly you've not touched on that, but what else enables you to kind of differentiate yourself in this market?

Speaker 2 7:46

So for us, and again PanZera, we're the mesh, we're the fabric that allows you to have, you know, a little bit of cash in your data center and everything else inside the cloud.

Speaker 2 7:59

And as you need to retrieve data, we use AI to make a smart guess at what you're going to need at the edge and have it there when you need it. So, for us, our ability to take help you with that multi-cloud strategy, enabling you to have part of your data in Wasabi as backup and some of it in Azure and AI, and then it's all at the edge. There are other tools that are tangential or augment the capabilities even that we provide, and, as a smart CTO, you're thinking about how can I squeeze the most value from the cloud? How can I enable my people to do their best work? Well, the first way to do that is to make sure they're safe, and a lot of the products or data transfer in and out of cloud isn't the safest right, and so you have ransomware attacks, you know, you have encryption, you might even have an exfiltration of data from inside out, and so having tools and platforms and again, penzer helps cover down on this, but there are others in the category as well who help cover down on just protecting your overall experience.

Speaker 2 9:18

Once you can sort of breathe and you feel confident that, wow, my people, my team, my data are safe, then I think your focus, evan it's a really good question you're asking your focus shifts to how do I become productive? You know what if, instead of it taking an hour to run a workload, I could do that in minutes? And then what does that do to my staffing model? What does that do to my optimization of services model? There's a lot of questions you can begin to ask and then, if you answer those questions, it's a matter of then just plugging in the right cloud. You might find that, you know, I can use a local data center cloud from a VAR. I can buy clouds from AWS or Azure or Google, and they can satisfy my basic requests. It's about understanding your long-term goal, or even your medium-term goal. What are you trying to accomplish in the near future? And then, as you move towards that, just finding a vendor that you believe has those options.

Speaker 2 10:32

You can squeeze pricing concessions out of any cloud. They all want to sell you lots of cloud, so you can squeeze pricing concessions out of them. But I think being taking the time to be sure you know what you want to achieve is that first step in you getting what you want. There's nothing worse out of than going through it. I'm sure you've been in these cycles where you spend nine months, 12 months, 15 months evaluating cloud, cloud services, cloud providers. You get the cloud, you lock in, you sign a contract, start moving data, and then you realize, well, this is wonderful cloud, but this isn't what we need, right? And so if you can start with what you need and where you want to go and then bolt on a cloud, I think you'll find that your success rate is much higher than everyone else around you.

Speaker 1 11:19

Wow, so well said. Well, let's drill down a little bit into the data security side of things. We know our CSO, ciso, cio friends are literally losing sleep over the challenges around ransomware and restoring and protecting data integrity. And what is your philosophy, or even your real approach here? Maybe give us a peek behind the curtain?

Speaker 2 11:47

Well used to say if you get hit by a ransomware, do these seven steps right, and it was all if, it was all if, and now it's when. It's now, it's when I think out of 400 plus customers. We've probably had 30 or 40 of them in a year. It hit with ransomware right.

Speaker 1 12:05

It hit with ransomware.

Speaker 2 12:06

So it's not about if it's when and the smart organizations right now are modeling what that would look like. You're gonna take your backups and ransomware attacks are getting sophisticated. They're timing that zero day attack. So it's right after you've taken a backup and it can do some corruption that causes you a low lecture, pain, or encryption that might cause you to actually pay the ransom instead of having access to a good backup.

Speaker 2 12:38

But there are some known things that are best practices. One of them is just having an immutable primary storage solution, not immutable secondary storage, but immutable primary storage. If you use a convolt or any other of the sort of world-class backup solutions that are out there, they'll all provide a secondary copy of your data that's immutable and in hours, days, weeks, months, quarters, years, you'll have versions of the software. But there are products that Panzera has and others in our category where your primary solution is also immutable and so as easy as it is. When you're in a Word document and you make a mistake and you're like, oh shoot, I deleted the wrong paragraph, Control-Z, and now it's back, I undo. It can be as simple to do in your file system or in file folders or files themselves that might have been corrupted.

Speaker 1 13:40

So one is just immutability.

Speaker 2 13:43

And the second one is thinking about the speed to restore, to restore. So it used to be in the old days of ransomware, and I'm speaking as if it's been years.

Speaker 2 13:55

but it's months and the old days all three months ago we would talk about like my data's ransomed and it's locked and I can't get it back and I don't know what to do and business operations are all falling apart. And in today's world we have backups and we're pressure testing our backups and we're sort of ready to restore, but restoring can take days. It's not easy to restore your files, and so again it sounds like.

Speaker 1 14:24

I'm pitching Panzera and I'm not.

Speaker 2 14:28

I'm just pitching a category.

Speaker 1 14:29

Please. We like pitching here.

Speaker 2 14:32

There's a category of products out there that ease your time to restoration, and again it's the ability to. Again, with an immutable structure, you can say, ok, I now know from an audit trail what has been changed by whom and I just roll that back. Add to that, the ability to detect any sort of malfeasance by anyone will help you reduce your blast radius. Here's what I mean by that the speed at which malware or ransomware moves. It's incredibly fast. It goes live and attaches to nodes and all of a sudden explodes and you might have 10,000 files a second being encrypted something like that. And the ability to detect when non-normal behavior is happening, sever the SMB connections that are causing this bad activity and interdicting a process that would normally corrupt millions of files in minutes. Now you've stopped it in 15 seconds. And, yes, you might have 150,000, 200,000 files you need to restore, but that's better than trying to restore 300 million files that might have been corrupted in a few minutes. And so smart CEOs are saying, ok, it's likely this is going to happen to me. Yes, we're testing our backups, yes, we're getting ready to restore, but what if I didn't have to restore so much? What if I only had to restore a smaller subset of data, because I was able to interdict, spot, detect and sort of obviate the risk that comes with bad actions. And, by the way, this is not just a ransomware attack.

Speaker 2 16:24

So much of the time we talk about encryption and data loss, but a lot of the unspoken damage that happened to the side of our organizations is exfiltration An employee leaves, they take terabytes of data, sometimes petabytes of data, with them. We know about these things, evan, when it comes to lawsuits, when Uber and Google. They get into a fight about who has auto driving technology from one engineer who left a company and went to another company. All of a sudden a similar technology that was at company A is now at company B, and then that calls for a multi-billion dollar lawsuit sort of shaking the legal scene.

Speaker 2 17:16

But in real life this happens thousands of times a day where employees are extracting data and leaking it, sometimes on purpose because they want to take radical action, other times because they just think it's important and they're going to put it on their laptop and that laptop might have some malware on it. Or they put it on a phone and that phone's been hacked or cracked or something's. It's leaky. And so the ability for CTO, cio, cisos to be able to spot non-normal behavior and then shine a light on. Is this something I need to interdict or is this just I need to coach up better performance from my employees? These are so powerful and just maintaining a safe, highly productive workforce.

Speaker 1 18:11

Wow, so well said One thing. You also said you're using AI today to enhance your cloud storage capability, data management capability. Maybe talk a little bit about that. It seems like you've been doing it before. It was a cool thing to do the last year, so tell us about that and what we can expect next in terms of AI, ml related operations.

Speaker 2 18:39

Yeah, so there's, our clients are enterprise, so these are the largest companies in the world doing some of the most complex work in the world, and not that it's nerdier or more important than what's happening in small boutique companies and Silicon Valley, but it's just the fact that these organizations are big and have tens or hundreds of thousands of employees across the world, across many locations, and so we've been applying machine learning to the 12 billion plus snapshots that we are managing at any one time.

Empowering Data and Collaboration for Innovation

Speaker 2 19:17

At any one time. We're managing some 12 billion plus snapshots. So, thinking about that, every day, 12 billion snapshots, and then tomorrow 12 billion snapshots, and this continues, continues and continues. And our theory of the model is that we should be able to know not just what is bad. And we spot bad because of current behaviors. You might encrypt the header of a PDF, right, or you might encrypt the file. That's bad behavior, it's known bad behavior, and there's 15 or 16 of these things that are clearing red lights, and then there's another 200 that are sort of orange lights that could be bad, and then there's a whole bunch of other behavior.

Speaker 2 19:59

We've been studying this problem from the opposite angle as well. Evident and I think this is the future of AI Instead of us looking at what is bad, why aren't we talking about anything that's not good? Because protecting our employees is just a baseline for doing any work at all. So if we don't protect our employees, they don't have access to their data systems, data tools, and so they can't do any work at all. They're basically looking at a machine that's fully encrypted and they have nothing they can do while all of our smart nerds are working to help them.

Speaker 2 20:42

So if we don't protect them, they have nothing. But what we've not begun to dream about this is where we asked our data scientists, our creative innovations that we think will help change the paradigm and provide some potentiality for senior data leaders is what if we were able to help shine a spotlight or anything that's not good? So we can analyze an organization and say here's what good looks like. A data engineer does these five things every day and a software engineer does these 13 things every day, and the sequence with this type of file, from this location at this time of the day, this is what good looks like.

Speaker 2 21:21

And anything that doesn't line up to it, either at a role level, location level, or even Dan Waltzman me as a person level, doesn't line up to that I can flag and then, as a smart leader, you just decide how many deviations from the norm before you flag it, stop it, counsel me, investigate or just start looking around it like man. How could we help Dan do his job better? So it's not just spotting bad things and shutting them down, it's not just noticing hey, dan's downloading too many things on this day. It's going beyond that to say, using AI man, when Dan's at this location, he works so much faster. Well, why is that? And then I noticed, oh, by the way, he has three times the bandwidth at that location.

Speaker 1 22:07

I wonder if I extended bandwidth to my other locations.

Speaker 2 22:10

How would that impact our workforce and how would we be more productive? And maybe that allows us to squeeze 3% off the bottom line, which is a $2.5 billion add back. And now, all of a sudden, an expenditure of hundreds of millions in bandwidth saves us billions of dollars in bottom line productivity. You can begin to build these models that are super exciting, but only if you empower your data landscape with a view that I don't think. Currently, evan, we're talking a lot about. We're talking a lot about beating the bad guys and stopping the bad guys, but what about the good guys, the hundreds of thousands of millions of workers who are knowledge workers trying to do their best work with data and tools that are insufficient often?

Speaker 1 22:57

That's so well said. Well, now you're really talking about team collaboration and communication across boundaries. You see so much. So many great US companies do a fantastic job of supporting their teams inside the US on our local providers, but not so much outside the US. Sometimes they've been second tier, third tier citizens in some ways in terms of collaboration and tools. So it sounds like that's a common theme. It's a global, distributed, geographically distributed network. Talk about some. I don't know if you can mention some of your amazing customers and brands and logos you work with and how they deployed you, not just in the US but globally.

Speaker 2 23:44

When you think about how organizations need to work together. You used the word collaboration a few times, evan, and distributed. Our distributed workforce is just it's where we are right now, and COVID made that distribution really more granular. Right, the edge is in a data center. It's not our office here in Greenville, the edge is right here in my house where I'm working right now. Right, that's my edge.

Speaker 2 24:14

And so there's always been this need to have everybody working off the same document, sharing one document, without the need for making three copies and two little colos and sinking it across the globe to follow the sun All of that model which is sort of the bedrock of our outdated data policy.

Speaker 2 24:40

There's been a need for something more, and so sophisticated users of PANZER are often just using our technology to allow to put a little bit of cash at an edge, which might be a data center and it might also just be my iPad, might just be my laptop right here, putting a little bit of a cash there and then everything else in the cloud and in something like Object Store where I can transfer those objects back and forth highly securely to the edge and serve it up to you right where you need it, which is next to usually the systems machinery that are pounding on that. What this means is, if you're doing HPC, you could have HPC in four different continents. Maybe a team in the US where engineers are at headquarters working on new designs, a team in India who's doing QA. You might even have a manufacturing facility in Germany, let's just say where they're building out sort of the prototypes, qa, and then you have cloud, because that's where we started this conversation doing all the high-quality computing.

Speaker 2 25:54

So I need to get data from the US QA to India, manufactured in Germany, and I have picked a zone where I'm now going to be doing the HPC in the cloud, and all of that with conventional systems outside of PADS are just not possible. But now, having this mesh where someone in India is viewing the same file but you're just streaming the blocks that aren't at the cache in India and any updates that India makes, maybe the blocks that have changed get transferred to the United States and vice versa to Germany, You've now got this web, this mesh, where you're quickly, quickly, able to iterate on changes. Everybody gets exactly what they need, just in time for their data purposes, and boy does it just accelerate time to innovation, time to value for their customers, allowing them to produce better products with higher quality, because making a product that reaches the world, what we all use and consume, if you have more time to QA, more time to pressure test, user acceptance, all of that, more time for those things, you get a better quality of product.

Speaker 1 27:18

Fantastic. That's so exciting. So tell us. I'm sure you have a very long and probably somewhat proprietary roadmap into the future, but what can you share with us on the future development, either on your end maybe a peek into your roadmap or on the Cloud side? What do you see as the most urgent, interesting development that you're excited for?

Speaker 2 27:43

There's a couple of things that I'm excited for. Some of that we're not actually working on. I'm just observing. For this year, there's two big themes for us, and it's stability and performance.

Speaker 2 27:56

Again our customers are doing the most complex work in the world. Stability is what they count on, foundationally rock solid, doesn't break and is smart enough to adapt to the needs of the customer. Performance is about speed. It's about integrity of data flow. Those are our themes. They will continue to be our themes. Some of the things I've seen out in the marketplace is we think about AI around, maybe the chat GPTs of the world. I start a sentence. It finishes a sentence. I talk to it it talks back.

Speaker 2 28:29

Those are neat. Some of the things that I see out there is just using AI to look at datasets and inform what's in a folder, pointing AI to folder and saying what is in this folder? Maybe using it for M&A Evan. It's like, hey, I'm buying this company. I'm obviously assuming liability for a lot of these things. How do I combine all these things together, all this knowledge together, into a risk assessment? Right, it's cool. I throw AI at a folder of documents and it returns to me those answers. Wow, I see some exciting things out there in the AI world, machine learning world in it. I'm excited to see how some of these things evolve. It does demonstrate the ability to help us do our jobs better.

Speaker 1 29:24

Wow, that's a great bottom line. Tell us, change is hard, it's painful. How do you recommend to a new potential client to get started? What's that initial step? Either internally or with people like Panzora doing things differently, not just doing what you've always done, which is the easiest path, of course, to take?

Speaker 2 29:53

This is probably the foundation of the question that was so I'm glad you're asking about this. I think it's just start the conversation, set the bar low. The bar does not need to be moving to the Cloud. The bar does not need to be. We're changing. Everyone loves changing management. What does it need to be that?

Speaker 2 30:13

I think the first step that I've seen be the most successful is starting the conversation. This is going to sound perhaps overly simplistic, but it's continuing the conversation. Once at the leader, you start the conversation, and certainly organizations like us, the bar that you trust, is another great option for you. The Cloud providers themselves also great for these conversations. Once you understand your optionality, what's possible, then you can begin to start enrolling other people inside your organization into the conversation. Maybe that the Chief Data Officer starts it. You bring a CTO, a CISO, maybe a head of support or professional services. You're bringing them into this conversation and then you can get other perspectives that inform what you ultimately need from the Cloud.

Speaker 2 31:09

I think one of the mistakes that I see it's a unnecessary collateral damages is you set the bar so high. We're moving to the Cloud. It's going to happen in six months and nothing's going to stop us. None of us know what we don't know. If you have the conversation, begin to think about optionality. All of a sudden, you're going to be able to have an enriched experience where you can then decide what's appropriate Speed and spend for going forward.

Speaker 1 31:44

Well, that was certainly a mic drop moments. I appreciate that. As we wrap up Last question here. You've been on the road pretty much everywhere all at once the last few weeks. What's next? Do you have a bit of a downtime or do you have anything else you're looking forward to over the next month or two?

Speaker 2 32:02

I'm excited by Panzerans stepping up and leading the industry in this discussion around both obviating the risks around cost and security and Cloud. I'm excited to see us grow. But look on a personal note, I love being home, spending time with the kids and just being a human within a pair of running shoes on the trail. So I'm excited.

Speaker 1 32:29

All right. Well, wonderful to speak with you. Thanks so much. As a thought leader, it's been incredibly informative and so insightful. Thanks everyone for watching Reach out and with any comments, questions, clarifications. I appreciate your time. Thanks, Dan. Thanks so much.

Speaker 2 32:46

Thanks, Alan.

Speaker 1 32:47

Take care everyone, Bye-bye.