Video: The Hidden Risk for Financial Services Webinar with AWS | Duration: 3008s | Summary: The Hidden Risk for Financial Services Webinar with AWS | Chapters: Introduction and Welcome (30.01s), S3 Data Protection (173.83s), Data Protection Strategies (388.13998s), Data Resilience Strategies (829.815s), Migration Data Protection (1359.925s), Financial Services Applications (1738.655s), Financial Services Evolution (1909.085s), Data Resiliency Strategy (2047.625s), Data Recovery Strategies (2230.6401s), Immutability and Recovery (2610.6602s), Closing Remarks (2657.2751s)
Transcript for "The Hidden Risk for Financial Services Webinar with AWS":
Hi, everyone. Welcome for today's webinar, the invisible thread between Amazon s three and business continuity. Joining me today is Matt Evina, a financial services specialist sales leader over at AWS. And I'm Adam Reschro, our director of cloud specialists here at Rubrik. Pretty excited for what we'll be talking about today. Matt, you wanna give a quick intro of yourself as well? Yeah. Yeah. Sure. Thanks, Adam. Yeah. I lead a portion of our financial services business for The US, some of the largest financial companies across the country. I cover core services, which include security and storage together. So this is a hot topic for all of the conversations I'm having with customers when it comes to their protection of AWS workloads. Or another big topic is, as people start to move workloads from on prem or other clouds to AWS, how do you build foundational and operational data protection in place from the beginnings, where we'd like to lead with customers as well. Great. Let's dig in. The topic of data protection for our customers is really big. I've read a couple industry reports lately that surprisingly, with all interest and heat around generative AI for the fourth year in a row, cybersecurity and resiliency risk was the top priority for CIOs. And they said a lot of it was driven both based on the emergence of even more sophisticated attacks, AI being used as part of the threats they see, and then just resiliency as a need for something that customers are building into their business. So, we're having more and more conversations with customers around the topic of data protection. And a lot of people think about it with emerging threats around cyber attacks, but there's foundational and operational resiliency practices that customers still can do a even better job of. So, Adam, really appreciate being able have a conversation with you today around this. Yeah, I agree too. And I think with that being a key initiative for CIOs, I mean, usually those, the top five reports, typically when you see them too, it's the top five, they're kinda close usually. Then there's just the one that's first. It's far and away on the cyber resiliency side too. It's a kind of distant second for the next ones on there. So just goes to show that this is still a critical need and something that businesses large and small are still trying to solve for today. This really matters to our customers at AWS because S3 is our biggest storage service by far, and it's built directly into our customers' critical financial services applications, whether it's things like their business and analytical data lake, core banking and payment systems for our capital markets firms. They're pulling in a lot of market data that they use to execute trades and trading strategies against. Some of our insurance customers are building out risk and analytics platforms that are built on top of s three. So just thinking about the proliferation of all the documents that get generated and shared with customers are foundational to S3, well as being in a highly regulated industry. Our customers are storing a lot of historical backup archival content that is required by regulators. So, S3 is just a massive storage service and also one of the big areas that customers, you know, are considering putting even more protection around as it relates to their core business applications within financial services. Yeah, I'd say too, S3 is probably one of the first services many organizations leverage, just even if they started down an AWS path years and years ago, where they're still mostly in the data center. But as their needs and their businesses have evolved, their use of S3 too, instead of just maybe that archive target or a backup storage target and things of that nature, dramatically shifted, think as well, which is probably where this is also even more top of mind for them. And how are they handling the needs for these newer applications, as well as some of the kind of original use cases S3 brought? Yeah, absolutely. And when we approach these customer conversations around protecting S3 data, there's a couple of different ways to think about it. One of the ways I approach it is talking to customers about how would they have traditionally There's the customers that are digital ID have been true born on the cloud. And there's the others that are moving workloads from on prem or maybe in a hybrid workload mentality. Those folks that are moving either completely out of the data center or moving applications to AWS. From a traditional on prem mindset, a lot of them are comfortable with the I need to have a prod and a Doctor. I likely need, you know, some local operational restore point. I probably have a backup appliance and then maybe even an offline copy Cyber Vault. But when they bring that data to AWS and and put it on s three, sometimes customers might mistake availability and durability of s three for built in protection mechanisms. So a lot of the conversations I have with them are like, how do you apply the same approach to the different tiers of data protection when it comes to workloads in the cloud? And you still wanna come to it with a different datasets have different value and criticality for restore points. And customers need to just like have a good handle on classification of their data and make sure that, again, as I said before, they're not mistaking availability and durability of S3 for all of my data is protected and I'm good. Agreed. We see that a lot too. I think one of the common aspects there too, is the point you hit on with data classification itself too, and really tearing out the importance of the varying data that may be in these S3 environments, because it's not a one size fits all solution at the end by any means. Yeah, absolutely. And we get into, think this is one of the points that customers will struggle with is sometimes as data grows and it always grows, but as capacity become much larger, it can be a little bit overwhelming for customers to think about how they have to protect data. And I've had a lot of different conversations with customers, whether they're insurers or in the data providers or capital market space, and like to make sure that they're approaching it from what's important to them and what would be meaningfully painful, both to their customers and from a financial perspective, if they weren't able to recover quickly. I think one of my customers said it best when I was saying, how do you all think about tiers of protection within your organization? And the customer said, they've built an internal framework around what is highly sensitive data, right? What could be included by PII or PCI data falls into a certain category of must have really good data protection mechanisms around it. The second was, clearly, if we're in a regulated industry, what are we required to hold in an immutable way that we could prove that we have a restore point from? And the third was, hey, even if we're not required to, if one of our services went down, either from an attack or operational outage issue, or some internal human error, what would be very painful for us from a financial perspective, either reputational damage to customers, or services offline, it means we're losing out on millions of dollars per hour, that falls into its own category of, we need a quick restore point and the ability to bring back up these services. A lot of customers have thankfully bought into this concept of like, what is the minimal viable company or minimal viable service? And, it helps feel break down when our CIOs and CSOs go talk to the board, they're saying, we're investing in building resiliency at this level, because we've classified what are the services that we require, absolutely have to have up for us to perform functionally as a business. And that has helped them get the investment dollars to make sure they've got the support to have the team's culture and processes and tools in place to really be running data protection the right way. That's great. That's great. And I think even some of the new regulations, overseas with DORA, has even reinforced a lot. It's good to see higher and higher adoption of this framework. Yeah, absolutely. Let's talk a little bit about data resilience itself too. I'd love to hear your point of view on that and kind of the varying ways this can be handled. Yeah. I think that data resilience of S3 is an important conversation with our customers because there's different ways that customers might approach it and feel like they're well covered versus more mature customers understand the difference between I have some foundational level of protection versus like, I can confidently restore and know that it will be a smooth and easy process. So, we talk about, you know, I use the example of the way people might've orchestrated a data protection scenario on prem. The way that we talk to our customers about within S3 protection in the cloud is you still wanna make sure you have something like a fast local restore point. It could be file level snapshots or versioning within S3 buckets. Some people operate in a environment where they either need true, beyond AWS single region resiliency, but need cross region replication, either based on where customers are accessing data from, or they need to be able to do quick fail over and run-in a multi region environment. So it could be cross region remote replication. We also talk to customers about making sure that you don't just rely on object versioning and snapshots as your restore point. Because with the billions of objects that customers have and and multiply that by the number of versions they might have per object, customers have told me too that while you might be able to do object level restore, if that was your single point of recalling from a past, either somebody messed something up or we had an incident and you wanna be able to restore confidence from a last known good gold copy, having a point in time backup that you can recover from, makes that very seamless and easy versus being able to go object version by version. And then there's the last part, which is really like, you know, company crown jewels. And a lot of people are starting to either because of regulations or understanding the criticality of their data, investing in cyber vaults or logically air vaulted environments, where they know they've got, account level segregation and locked down access over who can even pull back from backups in a clean way. Yeah, I think the common thread between all of this too goes back to that classification even as well too, because it comes down to the common of price and performance versus operational time, and putting all of your kind of solutions in one version of these without classification built on top of it, because it leads a lot of operational overhead as well too, that can be mitigated by those good classification strategies that I often see. But yeah, as that data landscape changes as well, and the criticality of that data is different for each S3 bucket that they may have in their environment. This all becomes something that builds on top of each other in and of itself too, where you do end up with that cyber vault on some areas, but you're still probably gonna have something for operational recovery as well for that same exact thing. If have a large bucket and you only need to roll back a couple of versions, you don't wanna bring the whole entire thing back with you either. Exactly. Exactly. And then as more of these vaulting conversations have come up and customers grapple with like, what should go in the vault? I like to kind of use a simple analogy of, right? Like, I live in a house, the house has lots of stuff, protection has costs. I'm not I don't live inside of a safe, but we do take the really important documents, family heirlooms, some small amounts of cash. I'm not gonna say how much, right? Goes into a a safe within my house that is much harder to access and is fireproof and all kinds of tamper proof, right? So, really wanna talk to customers about as you go through and profile of things for your business, what truly are your corporate crown jewels? And even putting ten, fifteen, 20% of that most important data in an approach that might have more costs, but really means that the things that you've invested in and help you stand out versus your competitors, are protected in a way that you know is truly immutable. You could restore in a way that has logical separation from the rest of your environment, and be able to restore and not reinfect everything else that you have going on. So that's the approach that I like to tell customers is our thought process on what you should be vaulting. It's not everything, but it is a very important portion of your business. Couldn't agree more. Let's talk about, I mean, what's the overall kind of data resilience strategy for S3 from AWS viewpoint? Yeah, I'd say our data resilience strategy is, it kinda goes back to customers having to apply their own sort of logic on what's most important. The best version of good data resilience strategy still falls back onto a three, two, one, one, zero approach, right? Making sure that you have three unique copies of your data, in addition to the original source. A lot of people are applying things like object level versioning or file level snapshots, to something you could quickly restore from. It might mean data that based on your either operational plan or regulatory requirements, being able to be in multi region the same time. So, it could be cross region replication and having data in separate accounts and separate regions. It certainly means for different really important data sets, having a point in time recovery that you can restore from that is policy based, and you know what intervals you would be able to go back and time travel through and restore from. And then having at least one copy. But you can turn on object lock for immutability at the object level, but having your most critical data put into some sort of isolated air gap vaulting environment is is essential. And then the zero comes in is which is like, if you do all of this, you also wanna make sure that you know that you can test it. So zero issues, right? Like putting all these systems in place without testing and knowing that you're gonna be able to restore without an issue, is maybe one of the most important parts of the approach too. Couldn't agree more too. And I think that's the one that tends to be missed. I think most often in my experience too, is that regular focus on just doing some of that testing. The worst thing you could do is test when you have an issue. That's the last time you wanna be testing anything. Having that, be it quarterly, monthly, just regular, some sort of testing to validate that, your overall data resilience strategy for those S3 environments each one of these levels, to just from even a version recovery or a disaster recovery, or even recovering back from that logically or got fault too, needs to be at all levels and really have that runbook established, written down as well too. So that when that moment does come, everybody knows what they have to do. Yeah. Adam, I share some of the kind of signals we've seen and see if you hear this from your customers at Rubrik too, which is just the the importance of having the cyber vaults and at least a point in time recovery backup copy is essential because the attackers we're finding are going after customers backup copies. Right? They understand that if you can get in, and what we found was like 90% of cyber attacks within financial services included in an attempt to go after a customer's backup copy. When these attackers really wanna inflict a lot of pain on our financial services customers, Going through and encrypting backups and locking customers out of restore points is one of the most damaging things that they can do. And unfortunately, we're seeing a lot more of those type of attacks being targeted at backups. And that customers found that they faced like a 8x higher recovery costs if backups were compromised versus when they weren't compromised. So, that immutability and really thought process into investing in protection of your backups meant a lot to our customers. Kind of like, it was a painful and expensive recovery process when they didn't have the right level of protection in place. Yeah, we see very, very similar. Typical entry point still tends to be predominantly identity compromise. And then once you have compromising the identity, you have a lot of lateral access to move around in that environment. Years and years and years ago, when we first started seeing cyber attacks, ransomware attacks, the backups weren't usually touched, but that was where somebody would still get a ransom note. There'd be like the encryption in place, and then the team would just recover from the backup. And then, you know, the issue was resolved, remediated. They were able to roll back. But the bad actors quickly caught on to that and said, well, if the way out of making this hard is to avoid making sure that extra backup copy no longer exists, we'll go after the backups. And that was really just a logical next step for a lot of these organizations to make it really, really impossible for these organizations to be able to safely recover. Because the end goal for most of these attacks is to make sure that ransom is ultimately paid. And if they're able to get to the backup data and no recovery is possible, then that's where a lot of organizations end up kind of having their back against the wall, which is highly unfortunate, but it's very common still to this day, unfortunately. I'll say one more thing and one of your take on this one too, which is- we talk to customers about the importance of having operational recovery and point in time backups. But they've also shared back with us, hey, if one of the challenges is not knowing that we have a good, clean, copy of data. Like, the importance of knowing which version of a backup to restore from really can save you time. Because it it it might not be the last version. If if you had an event, whether it was a human error or an attacker got through or anything happened, if you can't look at what that potential blast radius was and say, Go restore from this version because we know we're not gonna reinfect any systems, or we know that there's not gonna be any issues when you go to restore. That's something that customers have brought up as a key concern. Wanted to From your side, how do you approach that conversation? Yeah. That is really, really important. I think that's the difference between having what we've kind of tried to express to a lot of our customers too is the concepts of RTO and an RPO have been around for a while. RTO, how long is it gonna take me to recover? Since early data center days, that's been a definition that a lot of organizations have. If I have an outage due to a hurricane in a data center, I know I can fail over to my disaster recovery region, be back up and running in twenty four hours. That doesn't translate well to these cyber events at all. Because it's not just a matter of failing over. Number one, if there's a cloud account compromise, you might not be able to trust the identities on the other side. Number one, have to spend time just cleaning that up. But more importantly, that latest version is really meant for that operational resiliency where there's a bad script that might've deleted some things that you're rolling back from or a bad update. The CrowdStrike instance from last year was a good one where that was a massive operational recovery. You could go back to the latest version and be back up and running. But in these cyber events, typically the recovery mechanism is to restore into some sort of isolated environment, And then you're running a lot of forensics. And then if that's on the latest version and that latest version is still dirty, you have to rinse and repeat this process over and over and over again. And that's really what elongates the downtime here. You don't know which version you can go back to outside of any means, of trial and error. And trial and error is not an easy way to meet an RTO by any means. Especially when we're talking about the level of scale that many organizations are using S3 as well too. Simple domain controller or something like that, sitting on an EC2 instance, you could trial and error that in twenty four hours, you'll probably find the right copy. But with these large scale data sets, that's where we see a lot of the more organizations who have had really elongated downtime due to cyber incidents, talking weeks to months. It's usually because they can't find that clean point to recover from. And that just takes so much time. Right. That ties directly into the reasons that we're telling people, you need to have more than just operational recovery from a version or a snapshot. You have to have policy based point in time recovery backups enabled to be able to you know, have something you can run forensics against and know where you're good clean gold copy of your data is and then help your your business and your application teams be up and running as fast as possible. Couldn't agree more. I wanted to talk one one other thing was just customers who are who are migrating workloads and moving fast. This is a pattern that we see is also like sometimes when the use cases are moving to the cloud, there could be a change in ownership of who are responsible for data protection. Maybe in an on prem world, there could have been a team directly responsible for backups, but the roles and responsibilities of who's in charge of infrastructure and data protection, as some of these use cases move to the cloud, I've seen has shifted. And, I also recommend whenever I'm talking to customers, they've got data protection built into migration plans. Because sometimes people might have a top down directive to move an application quickly, or they've got a defined timeline and they might not have built into appropriate planning or ownership or processes of how that data will be protected once it moves to S3. And I said before, there's a difference between us serving up extremely high availability and durability. And, when you move workloads over to a cloud environment that are outside of your data center or a different line of control, customers need to think through, you know, who is that? Is it infrastructure? Is there a backup team? Is it on the application or line of business themselves to go back this data up? But, this should be something that is built into a migration plan, is one of the recommendations I talk to customers about all the time. And, as my team is responsible for not just security, but storage, we're often on the front end of people thinking through data migrations to AWS. And, we're helping them say, All right, here's the right recommended storage service. As you're moving these workloads to us, are you building an operational, foundational, and cyber level protection? So, do you see the same thing when you talk to customers who are moving workloads to? I do. And I think especially for those migration strategies, because you're right, the general kinda, I think overall roles and responsibilities and personas dramatically changed too. Especially since I'd say anybody who's doing a migration, more often than not, they already started AWS doing something prior to the migration itself. So it was more of starting to build new apps in there and there's already a team focused on that. And then in parallel, there's a lot of migration efforts, be it to get out of the data center or just move more critical workloads. And as the new people and processes end up in place too, a lot of them might not have the background previously, even in the data center where they're constantly thinking about things that fail all the time. There's just lots that go wrong in a data center on any given day of the week. Fiber lines being cut, hard drives going down, power outages, you name it, where all of that is top of mind. You move to a service like AWS where just out of the box, you get a lot of resilience and a lot of things. And that's one of the major selling points of building on AWS and building your applications for all these failure scenarios in mind. But that's usually where there's that kinda push and pull and just the knowledge that I think is there for some of these teams where having that overall end to end resilience strategy built into the application as soon as possible is really critical. Especially for a migration scenario. This is usually a lot of data that's moving a lot of critical applications. If you do it after the fact, it becomes a lot harder as well. Dramatically harder. And we've I've been in a lot of conversations where somebody will have a defined time label that they have to hit, and then operational resilience and data protection can become something that has to be addressed six to twelve months later after the workloads might have moved and already been put into production. And, that can leave yourself with a lot of gaps. So, you know, that's why I say whenever we're earlier on in migration planning conversations, we're starting to recommend, you know, what's the criticality of your dataset, as we talked about before? How important is this? Is it easily recreatable market data that you can pull in from a data provider? And maybe you don't need to apply a crazy amount of protection on, but if this is something that is either essential regulatory required data, sensitive customer data, or, you know, just downright really important to your business uptime, as you move these workloads to AWS, build in the right protection mechanisms from the beginning, rather than go try to clean it up, six to twelve months later. Yeah. I think even related to that too, one of the other unfortunate things I have seen when it is more of a, we'll get to that later. The later always gets lower and lower on priority because other new things keep coming in and coming in and coming in. And it goes back to that somewhat fault sense of security where it's like, my data center, I had an outage every other day and well, we did this whole migration and everything's been up and all the migration work great. Everything's live in production. So I can lower the priority of this. And that's the other thing I try to caution a lot against as well too. And if you can't get it in early, if you get it in early, you don't have to think about it later. And even if there is an issue during that migration too, having a backup copy as well during that migration means you have something to recover from as part of that. Exactly. And I think that's one of the important considerations too, like customers move these workloads and we're seeing some of the people wanting to move fast as they migrate that new VMware workloads. It's also through warm handoff between different teams that might be responsible for this, but also customers are very interested in how do they think through protecting the workloads end to end. Maybe I didn't cut over. Maybe I'm under the gun to get VMware services moved to AWS really quickly, so I can re platform and then re host them eventually. But, I'm going really fast. I need to build in data protection related to some of these big drivers, like compelling events around VMware right away. Yeah. I think that is one of those things that are accelerating a lot of this for customers overall, and just their data migration strategy. And then wanting to get to that move of refactoring sooner rather than later as well too. But once again, having that baked in early into the overall strategy saves a lot of time and pain later on typically. Absolutely. I'd love to hear from you too, Matt, a little bit more about what are some of the other common use cases and applications you see specifically that the financial services customers you talk to regularly are using Amazon S3 for, especially as it kind of runs the gamut of kind of that more traditional archive use case for those compliance centric and regulatory workloads to newer things they may be building. Yeah. I'm glad I cover financial services because I find the use cases really exciting. Like, the pace of change within financial services is sometimes people think about it as like older and stale and less. There companies are are reinventing themselves, and also, there's, like, a lot of reemergence over what, you know, what is the definition of a bank? So, with crypto and capital markets, everything is merging and changing very quickly. So some of the key use cases we see where customers are using s three for essential parts of their financial services businesses, the common ones you would think of around a business data lake. Right? People have tens or hundreds of petabytes of data where they've got different information and data points about their customers. This could be things that they've organically built up over decades of relationships with customers, or it could be if you're trying to do market analysis or understanding what's the right product or service to offer your customers with, where you're pulling in a lot of different feeds from different data providers that make that data lake really big. People are using the S3 as part of their analytics pipeline. So anyone that wants to do generative AI projects, you need a lot of data to train your models against. And, it's even better if it's your own personalized corporate data versus generic data. So, people want to be able to aggregate and store and hold onto analytics data lakes for generative AI projects more and more. We see customers within the payments and banking systems where S3 is tied directly in their core banking and payment systems. There's tons of transactions that require a lot of reporting, statements being generated and served out through a different customer facing application for all those financial transactions that take place. Our capital markets customers ingest hoards of data for them to be able to learn about back testing and trading strategies, being able to generate alpha in different ways from their competitors, through having a massive trove of data they've collected and analyzed and applied their secret trading sauce to. And then, there's just a ton of financial documents and statements that get created and sent and shared with customers are all tied into it. The other one is, whether you're in a capital markets or an insurance business, a lot of risk modeling and fraud prevention are core applications that we see customers using for Esther as well. Yeah, it's definitely evolved significantly as well too. And I've had the privilege working with various financial service customers as well too. And there is a lot that they're doing. I'm always surprised and impressed too, the things that they're thinking of. I think even the prevalence too, of just how many transactions they're processing in any given day, via the varying use cases, it's pretty incredible at how large that ends up scaling. Especially, I mean, just look how quickly it is for anybody in the world, not to buy anything anywhere. But everybody's doing it on their phone. I can't remember the last time I wrote a physical check, as an example there. And any ones that have ever been sent to me, I use my phone to process those as well too. A 100%. People have people have digitized and automated the whole mortgage and loan origination process. They cut weeks off of people going through and trying to apply for credit. So S3 is behind all of it, whether it's at the front end of a loan origination or the back end of somebody using digital payment application to send money to a friend so they can go and try to be the next mega millions or powerful winner. I think all of this is pretty critical back to what we've originally talked to as well, because there's a lot of business critical data sitting in these S3 buckets. There's a lot of sensitive information in there as well. And there's a lot of IP related data in there, and it's just all varying shapes and sizes. And the amount of data growing in these S three environments for these types of organizations just continues to grow exponentially. Applications that started in the tens of terabytes. I'm sure most of those are all petabyte range pretty quickly for a lot of these organizations and growing just the amount of data and scale overall. And I think with all of this too, that goes back to a lot of that classification we talked about. So you have some things that have that sensitive data in it. You have some of them that may be some of those more tech forward leaning where it's a little bit more of a science project, so to speak. This gonna work? Is it not gonna work? And then as soon as it does work, they gotta meet the market. So then all of a sudden it's production, potentially cross costing millions of dollars on a given day. Exactly. And these are all ultimately attack factors too, I think we see specifically for a lot of the newer cyber threats we see. And just tying that back even to AI as well. And these bad actors as well are just innovating their own pace just as quick as the organizations out there. So as the organizations are expanding into things like AI and ML to beat their competitors, The bad guys are trying to get just as smart as well. And this is a big area where having a true data resiliency strategy for these critical S3 workloads is huge overall. And with all of the pace of change in these environments, that's where I think it's just more critical than ever for all these different business units within each one of these organizations to really have a true strategy. And there should definitely be one at the overall enterprise level across that organization. But each individual sub component, each business unit needs to adopt a lot of these standards themselves. Even if it is just basic data classification and the different resiliency bounds that are on there, it really helps them prevent some of these bad actors, not only getting into their environment, making sure their data is secure, but when it inevitably, if there is any sort of inevitable issue where these S3 buckets are a focus of an attack, they're able to still recover from whatever scenario gets thrown away. Especially with the scale of S3, that downtime just elongates, I think, as we talked about earlier and having to try to restore petabytes of data and then analyze it and then figure out where that clean point is. It can take a really, really, really, really long time for a lot of these organizations. Yeah, Adam, I just wanna jump in. One thing you said there, which was in the past year, I was talking to an asset management firm and asking them about their approach to data protection, operational resiliency. And they said exactly what you did, which was, there's a lot of different ways to approach this. And, while some may think that things like the quick local operational recovery of object versioning by itself may have some level of protection, That person's like, I don't wanna be on the hook for bringing our business back up, then going version by version with hundreds of billions of objects across our business. If somebody said, event happened or, you know, we're worried that some kind of data was deleted or compromised, go back and help me be up as fast as possible. That customer's like, there's zero chance I'm gonna put my team through the pain of even signing ourselves up to go version by version to restores. It was essential for us to invest in something that had points in time policy based recovery that meant we could The event happens, human error happens, we're back up in minutes or hours versus weeks of relying on just snapshots or versioning on its own. Yeah. The amount of man hours to try to restore from versions with that many is unfathomable in many ways. I even like to think S3 now with just the way applications are using it. It's just a massive database in many ways. It acts a lot like databases. And if you want to talk to any traditional DBA in the world and said, Hey, do you wanna just piece all this back together one transaction at a time? They think you were crazy. It's very similar just with the scale of the S3 brings to organizations. And I think that point in time is really key. I think that's where just a lot of the the newer innovation that we've had with Rubrik on on this as well too is really where we're able to help a lot of the these customers achieve those those outcomes. So one of the things that we can help with right out of the gate in that data classification is really with doing some data discovery across these S3 environments and seeing what their current state is. Meaning if they have some of those tier one workloads and some that they've already classified as maybe tier two, we can look into those and see what their level of risk is for sensitive data and what they're doing today for data resilience and protection on that. So if they only do have versioning on those, but we see that there's some more sensitive data and or compliance related data, that's where we can make that recommendation of, you probably should do a little bit more here. Maybe that is the area where you wanna potentially put that in some sort of more vaulted environment and really hide some of the overall blindness that may be there. And really helping with the overall observability in these large S3 estates. Additionally too, I mean, the risk of data deletion, because Rubrik does provide that point in time copy, does give these organizations a quick point to recover too. And with flexibility too, we do see some organizations even leveraging Rubrik to do help with some of the data disaster recovery use cases as well, where you're taking a backup of that S3 bucket and you might have the need to have it in two copies in two different places at the same time for regulatory reasons, going back to 03/2000 model there as well. That's something we can help customers achieve with some of the efficiencies we have as well, where based on the dataset and the use case, we're happy to tier down lower storage class as well to help make this not a large financial burden to make sure that they still do have this more secure outcome for their critical data. And I think one of the biggest ones though, is the overall data remediation. We've talked a lot about not only protecting the data, but the recovery aspect as well. With a lot of our cybersecurity tooling that we have in place, we're actually able to, while we're doing just as part of that basic backup that we're doing, not only can we put that backup data in that secure vault, but as we're putting it there, we're able to identify anomalies that may be happening. That may be indication of things like encryption events across the S3 buckets. But more importantly, we can actually leverage our own threat intelligence that's baked directly into the product that can identify potential instances of malware or other malicious files that may be in these large S3 environments. So that when that remediation or recovery does need to happen, the meantime to recovery is dramatically reduced. We can do live pre hash scans of everything that we backed up. So we can quickly quarantine even those bad points in time, make sure that there's no just accidental reinfection into the overall broader organization and dramatically reduce what we call that cyber RTO. So instead of doing that trial and error of multiple resource and a lot of forensics analysis, we can help our customers quickly find that clean point of recovery and treat this large scale potentially cyber event more like a traditional operational recovery where it's a regional outage that they're recovering from. And that also means that the overall regular testing that they're doing, they're using the same tool regardless of the way that their data may have been impacted. So they're able to test not only basic operational recovery, but have more stick time as well too on these cyber recovery type of events. Yeah. I think that's great. When you you talk about like data discovery and and one of the biggest challenges is customers kind of understanding their own risks and gaps and what has to be classified and protected in different levels can can help them both save a ton of time upfront on trying to figure out where to invest and where their risks are for for regulators, but also, you know, the policies in place. You're not waiting six, twelve months to go through a science project of, yeah, let me go application by application to interview a ton of people, turn on five to 10 different tools to try to understand this. But be able to do that out of the same system that can apply the protection mechanisms is a big benefit for customers when they're thinking about this. I agree there too. I mean, six months of back and forth, that's not something the bad guys are gonna wait for by any means. And even some of the simple things with data lakes that we talked about a lot as well, identifying what the reproducibility of the data is, that's another challenge as well. And is it easier for me to back it up from this RDS environment? Or what's the recreation time? Can I recreate it at all? Those are additional things we can help to make sure that the right data protection strategy is applied to the right workload. Yeah. So something else you said was, I think really important, Adam, was I was recently talking to somebody that's part of a incident response team. Like, I can't imagine, you know, how much coffee those folks drink when they're they're called in to deal with something and how many hours they can go without sleep for. But but they were talking about when they're called in, part of their job is to help people do that forensics. So knowing that there was good clean gold copy that was immutable is really important, but they also want to be able to do that, like kind of isolated forensics to to track back how did this happen? How do we prevent it from happening again? And what they what they told customers is, you know, when they've had a something go bump or go bad, one of the biggest recommendations they told them was, hey, you can invest in a ton of tools, but protection at the data layer itself is like your last line of defense. So for people that might not have, and you're gonna do these after action reports after an incident, if people didn't have immutability and points in time recovery in place, that was one of the number one things that these folks on the incident response team said was, Hey, make sure that this data has a good clean gold copy you can restore from. Know what that version is. Make sure it's immutable. So you can be back up and running. The infected versions or something where an anomalous activity happened and run our forensics against it. But let us go do that. Take a couple weeks, but we want your business up and running much faster next time. Yeah. It makes that whole scenario a lot less stressful. As long as that end to end is in place there. Really can agree more with you. With that, Matt, I think we're pretty much at time here. Thank you so much. Is there anything else you wanted to hit on before we close out today? No, the final comment I would just make is like, we're a firm believer in customers building both operational and data protection resiliency from the beginning. So, anyone that's interested in either protecting workloads where you feel like you've got a risk, or you're considering moving the next set of applications, or doing a data center exit, let's go have these conversations as early as possible. Security is job zero at AWS. So everyone, no matter what your role it takes us extremely seriously. Anyone in the account team is invested in making sure you have access to the right resources to apply the right protection frameworks. Appreciate having a strong partner like you guys, who can come in and help cover our customers, both workloads that are already on AWS or people that might be using Rubrik on prem and are considering you know, that that hybrid level of protection and using AWS as a target. Like, we're we're excited to partner with you and we're even more excited to help our customers feel confident they can recover quickly from any type of event. Couldn't agree more. I mean, we we have a lot of joint customers already. We have, you know, a lot of joint conversations as well too. So it's always great to see that as we partner together, we're really helping to trust these customers, address some of the most critical challenges that they're facing today. Yeah. And thanks for inviting me to join you, Adam. Great to see you. No problem. Always good to chat with you, Matt. And thank you so much for the time. Really appreciate it. All right, see you. Thank you again, Matt, and thank you all who have joined today. If you'd like to learn more, please check the resources tab, and have a great day.