Name: Where Your AI Budget Should (and Shouldn’t) Go
Uploaded: 2025-09-23T16:15:13.019Z
Duration: 1 h 14 min 24 s
Description: Where Your AI Budget Should (and Shouldn’t) Go

Transcript for "Where Your AI Budget Should (and Shouldn’t) Go": Hi, everybody. Thanks for joining us today for our Nasuni Unified session entitled where your AI budget should and maybe more importantly shouldn't go. My name is Lance Shaw. I'm the director of product marketing here at Nasuni. And, it's my pleasure to be joined by Karen Price from Frost and Sullivan. Karen, can you introduce yourself to our crowd today? Thanks, Lance. I'm Karen Price. I'm industry director and AI program manager here at Frost and Sullivan, where I cover cloud and AI research. I've been here for twelve years, and I come with a background in industry covering network, data centers, and cloud computing. So happy to be here today to talk about AI and, how folks can be successful. Absolutely. I'm glad to have you here. This is, kind of a hot topic. Turns out AI is somewhat popular. I don't I just just recently heard about it. It's, Just a little bit. Just a little bit. Top of everyone's mind and and that's why we're here today. So our goals for the session today are really are to help you sort of understand the full cost spectrum of an AI implementation, what's all involved, and and and what to look out for, and really identify where your budget allocation should go to really get maximum impact. We are gonna spend a little time focusing on the value of corporate unstructured data as a component of that and where the that how it plays in that allocation model. And, of course, we're also gonna talk about what you can do to be long term successful long term, I should say, in terms of, you know, scaling profitably and man maintaining focus. So we're gonna dive into that and, we'll get started here. You know, let's start with a couple stats. You know, Karen, you you brought some interesting stats to the table. Maybe you can tell us more about those. Yeah. Absolutely, Lance. When we, we annually survey, close to 2,000 businesses on their cloud AI, usage, perceptions, etcetera. And when we did that in 2024, we found that 77% of businesses found that data security, privacy, governance is a challenge to their AI deployments. So they're finding data to be a big problem when they're trying to deploy AI and and do AI initiatives. We also found that 71% found that obtaining, managing, their datasets and whatnot, for AI is a major challenge. So, clearly, the data is one of the biggest issues facing our customers today as they try to handle their AI deployments. We also know that 87% of AI projects never actually reach production, which is a rather staggering statistic given the importance in of AI in business today. Yeah. That that's pretty amazing. I mean, I think go we're gonna delve into what some of those reasons are. And certainly, you know, we we we hear about that in the news even sometimes about how we're you know, these early initiatives are are failing and there's some good reasons for that. And one of the reasons that everyone's here in attendance today is to be able to make sure that you are not part of that statistic. So as the c suite increasingly recognizes the potential of AI to enhance business functions and drive growth, they're increasingly interested in AI initiatives within their organizations. Though senior IT leaders are still the driving force behind AI adoption in many organizations, executives are quickly growing their influence over AI deployments in the enterprise. As you can see, Frost and Sullivan found that 37% of organizations rely on c level executives to define AI goals, while 43% involve senior IT leaders. This shift reflects a growing recognition that AI impacts customer service, operations, and overall strategy, not just technology. Lance, are you seeing the same with the customers that you work with? A 100%. There are more people involved from business units in in different parts of the organization that are looking to realize, the benefits of AI in their particular, line of business. Right? So, so it's it's widening the number of people that are involved in those decisions. Yeah. It's really interesting. And yet, despite the fact that there's a growing or a growing recognition of the influence on AI in the organization, The majority of, enterprises, we find about 81%, remain in the exploratory stages of AI deployment. There's only 1% that have you achieved ubiquitous AI deployment, and that indicates that there's really significant room for growth. These numbers have actually held steady for the last two to three years, indicating that while AI experimentation and even some leveraging in specific departments is happening, few organizations are strategic and enterprise wide in their AI deployments. This really underscores a need for some foundational investment in strategy, in data, in governments to move beyond that experimentation level. Let's take a look at what truly mature business AI entails. So, you know, at Frost, when we look at AI maturity, we look at where companies are in a four stage journey across four key parameters. We've got strategy and roadmap articulation, data readiness, regulatory compliance, and policy implementation, as well as technology implementation. And in order to achieve full maturity, how businesses store, manage, and protect their data is a major element to consider. And that's why, I was really eager to get on the, on this call with or this webinar with Lance today in order to really discuss the data. You know, foundational elements like a unified data architecture and governance and infrastructure are essential. And in order to achieve maturity, we really encourage businesses to unify that business data, including that from real time sources in order to make it accessible to AI. Lance, you know, what are you seeing in the businesses that you, are working with in terms of where they are on that data spectrum and and what they're looking to do? Yeah. You know, in terms of maturity, I think it's it's happening slowly. I mean, if you if beyond what you might read in the press, I mean, just looking at the average organization, people are just still getting familiar with what they can use AI for. And and and that so that planning element is still happening. Right? So, and and I think there's a lot of testing going on, a lot of experimentation, which would lead to this high number of quote, unquote failures. The people are still trying to figure out how they're gonna use it. And the the AI, environment is changing rapidly. Right? There's a there's a new model every every few minutes it seems. Right? So, trying to figure out what is the right technology to apply to an environment and to a particular business process, it it it takes some time to get that nailed down. So it it when you talk about how, you know, everyone's up at 81% still in the exploratory stage, it's somewhat shocking that that hasn't moved. Initially, for me, it was anyway. And, you know, but then you think about, well, where people are going, they're just now starting to get some traction. So, I mean, I think that number is is absolute that whole chart's gonna change and eventually probably shift the other way. But it that's where we are. That's where we are today. I agree. But there there's a big education element, which is part of the reason why we're doing what we're doing today. Right? And and that we need to shift businesses into a place where they understand that doing this at an enterprise level is a really critical thing. Breaking down those silos, departments working together to make that AI work across the company and to ensure that enterprise goals are being met, right, not just one department's, little AI deployment. We want this to work on a company wide level to drive the best results. Yeah. Absolutely. And I think I think just another comment, Karen, you made me think of is the that organizations are, you're you're looking at sort of re you're rethinking your your strategy. You're looking to do, for some organizations, yet another digital transformation. Right? The the the, that that term has been bantered about for decades. The, I think that and so that there's some there's some leeriness of doing so. And and and you have to really see the value of, for example, unifying your your your stack, unifying your your data and its access within a single environment. You you you have to understand how you're gonna use the data, have plan that out first, and then then make sure that you make that strategic investment, to be able to do that. Yeah. I would agree. I we see that in order to achieve maturity, enterprises really need a plan. Right? If they create what we like to call an AI road map, they can define the goals, the outcomes that are expected of their AI deployments. And companies that do that well look at it at a corporate level, not just at those specific business units. And by defining those goals, it ensures alignment with the broader business objectives. Right? Enabling Yeah. They have to enable prioritization. They have to, plan for resource allocation, and they have to be able to measure their outcomes. Right? I mean, when AI strategy is limited like that to those individual projects or business units, They risk fragmentation, duplicated efforts, and really, when it comes down to it, missed opportunities for scale and for impact. Yeah. Yeah. And yeah. And then by the way, if your competitors are are are are ahead of you and and are ahead of on in in this game, then you're starting to fall behind. You know what, Karen? It comes back to the old adage. Right? If you fail to plan right? That's what this is what it all comes back to. What you we I completely agree. And today, we have found that only 27% of enterprises have defined their AI goals at the enterprise and project levels. It's crazy. So yeah. With the and with data being such a critical part of that, let's look at some of the specific data trends. Right? We all know that AI is only as good as the data it learns from. Right now, 43% of enterprises are struggling to establish some sort of unified truly unified data foundation. And we know that fragmented and siloed data leads to poor model performance. It leads to unreliable or even inaccurate insights. This is where folks like Nasuni and the global file system come in. They're able to consolidate data across the cloud, the edge, and premises environments and make it really accessible, secure, and AI ready. So if you wanna scale AI, you need to start by fixing your data infrastructure. Lance, tell us a little bit more about the potential impacts of not preparing that data for AI and what businesses really need to consider as they make that journey towards data readiness. Yeah. Thanks, Karen. I mean, there's a there's a whole series of of cost impacts that you need to consider and, you know, it's it's like the proverbial iceberg. You you are there are some that are very obvious that that folks say, oh, yeah. Well, I've gotta get my software licenses together. I've gotta do any initial model training or refinement that is necessary. I've gotta get a basic infrastructure set up. But the that might be just, you know, 20% or 30% of the cost. There's lots of hidden costs for organizations to consider. First of all, is the is that data infrastructure. Right? It's we we were just you were just talking about. Is you is do you have, really the the the data pipeline built and the necessary storage infrastructure to be able to deliver data, selectively with privacy and security concerns so that you're not causing yourself some serious headache downstream in terms of, regulatory concerns, compliance, you know, obligations that you have to meet. And then it's just really that this isn't a you know, this is back to the strategic investment idea. You there is going to be long term operational overhead. The overhead that you have to account for. Right? The just the monitoring of your environment, the performance, the maintenance ongoing, the the models are always changing. You're gonna what you're using today is not what you'll be using tomorrow. There'll be a new and better model in place that you'll wanna take advantage of. So there's this constant, sort of, you know, upgrade or flywheel that you that you're gonna be running to just maintain and and continue to develop and improve your environment. And there's cost there. Right? The other thing is gets into integration complexity. The more we use agents and the more that we're leveraging data from multiple systems for within a a defined process that we're looking to enhance with AI, you're gonna involve other systems. Some of those are legacy. Some of them may be incompatible. You know, they just they're they weren't built for AI. Right? And that kinda that can from an unstructured data perspective, that certainly gets back to, old legacy NAS systems, for example, that just really can't handle it. Right? And they cannot deliver then. They're not built for it. So there's gonna be some integration work. There might be some APIs. Now there are things like MCP that are emerging as standards that will help with that and minimize that, but because you don't wanna build single integration points for every single thing, that gets ridiculous. But, these are all costs that you have to factor in and plan for. Absolutely. Yeah. I I would absolutely agree with that integration complexity. We look a lot at Frost at hybrid cloud environments and the reality that has become hybrid cloud. You know, no no business is going to just get rid of legacy for the sake of getting rid of legacy and making this huge digital transformation at once. Nobody has the overhead to do that. Nobody wants to put out that capital expense. It it's not a realistic, thing for companies to do. So we take a a good look at, you know, the hybrid cloud storage, the need to be able to unify that data while leveraging existing resources, right, to be able to, do data, for lack of a better way to say it, in a way that enables the AI without getting rid of all your existing investments. So it's really critical to be able to do that in a hybrid way, have, the right management plan and control layer to be able to do that in a way that is unified, but also safe, secure, and easily governed. Yeah. Yeah. I think I I I agree with you that there's there's probably a a a phased or staged approach to rolling these out some systems. And that really comes down to your particular workflow and process that you want to improve. Right? AI is basically useless unless it delivers value. Otherwise, who cares? It's just another technology. Right? But if it can automate and improve and make things faster and smarter and that can get way more done, then fabulous. Right? And and so that usually, just because of budget concerns, you do you take one bite at a time. Right. For many organizations, that is, moving moving unstructured data into a single location. Even that can take time as a priority effort. You can move it in stages if because you can't do it all at once. We have too much data. So you do it you do it in sections and you do it in a way where your high priority data becomes available. And then when you finally have a when you've gotten to, I should say, sort of a a global approach to how you are going to deliver data and get rid of those data silos. I mean, everywhere you look, when you're talking about serving up a data foundation, data silos are the enemy. Having things scattered to the four winds, lots of different storage areas, you are going to get conflicting and or erroneous results that will that will deliver hallucinations to your users, which is absolutely not your goal. Right? So, it's it's really important to be able as you think about your plan, consider the the proper foundation for for your unstructured data. And that, generally speaking, is a is a unified name space. Right? Everything all your files in a single location. One place to go. A single source of truth for AI inference. You you gotta be able to do, you gotta be able to have that so you don't have those conflicts. You don't have different versions of the same story. You know, and eliminating data silos, is absolutely a critical part of that. And then the other factor that that comes into a hybrid cloud storage model as you described is delivering global file access quickly. I think people sometimes think, you know, if and and rightly so, depending on the the topology of your infrastructure, if you've got a lot of, you know, remote locations and they are contributing or need to access data and use it in conjunction with AI processes, that needs to be relatively quick. That needs to be accessible. It It can't be, you know, please hold and look at a spinning wheel. People don't want that latency. Right? And so that has to be eliminated. And that's that's actually one of the values of a hybrid model is that there are access points at the edge that deliver that data really quickly. So you feel like you're right you're right on the still on a land. Right? And let's face it, Lance. For many, industry verticals, that's really important. You know, look at, for example, customer service where you've got remote agents working from home, remote call centers all over the globe. They need all to all have the same access at the same speed. Otherwise, you're not just risking, incorrect data, incorrect information, which is a big thing, but you're risking, poor customer experience. You're risking the reputation of your company. So that's where we get back to that c level influence. And, hey, these aren't just things that are, making a small problem for productivity or for a single department. You're risking corporate wide problem, reputation, loss of revenue. If you don't have, the data that you need accessible in the location that you need it at the time that it's required. So Yeah. No. Absolutely. And, you know, the other factor of that is sometimes that the, you're gathering data at those edge locations. Right? Maybe you you mentioned verticals, like in a manufacturing sense. Maybe you're taking data right off the line in a factory in another country. That needs to get incorporated and made available to AI immediately as part as to for the rest for the benefit of the rest of the organization. If that takes a very long time to get there, then you're you're losing hours or even days of productivity. So you wanna be able to get that quickly and having that edge access is is pretty critical for that. Absolutely. Yeah. I I think, let's let's talk about what happens when you bypass a data strategy. You might have some comments on this one. We were joking about the the GIGO acronym. What is GIGO? It's the world famous garbage in garbage out. That and that happens with if you don't have a data strategy and you're like, we're gonna build a bunch of connections to all these different silos and it'll be fine. We'll just go with what we've got. We don't wanna spend the money or the time. Like, okay. You're you're gonna you're you will suffer the consequences, and you will you will have, you you will suffer from a lack of data quality. Because, really, this is all about data quality, and accuracy. Yeah. That that's right. We in in our work with clients, we see garbage in, garbage out frequently. Like, businesses may start to dabble in AI, primarily using, like, open source models that aren't trained specifically on their corporate data. And, outcomes are not what was expected or they have limited applicability. And, you know, part of this, I think, is a governance issue. They haven't tackled the data protection and governance. There's, the challenge of intellectual property. And so sometimes I think that, engaging with a trusted provider to help you on the journey can be a really important thing to help you tackle that data that data issue. I know there's also legal risks. You know, you might wanna speak a little more to, other aspects of bypassing that strategy. Yeah. I think, you know, from a legal side, honestly, I am with the with with the rise of agents and people starting to play with agents and they're they're and connect to multiple systems and and really, really start to drive process, really start to make a difference. I'm waiting for that big company to be exposed. We've already seen some cases where data got wiped out because an agent thought it was the right thing to do and just wiped out an entire database and shut the company down for a while. Nobody nobody on this call wants to be that company. Right? And so that gets back to planning and having a data strategy and what data is exposed and what these agents are doing. And we'll talk a little bit about resilience here in a minute. That's another huge part of this story. And so, you know but we talk about not having a data strategy. There's obviously when you have duplicate storage, you're you're you're spending money on stuff that you really don't need, which is never a great idea. So you're you have infrastructure waste. The other thing is when you don't have a plan and you don't have a strategy for access, people will go off and do their own thing, it turns out. Right? So it's Shadow IT all over again. In this case, it's Shadow AI. People are are running models locally. They're bypassing security. They're bringing in their iPad, to work and connecting and doing things that they shouldn't be doing. All under the auspices of being more productive, but they're introducing a ton of risk. Right? So you can't have that. Absolutely. Absolutely. Yeah. Let's let's talk a little bit about security. You know, the and and security is really you know, we we we talked about it here as, the non negotiable investment. This is something you do not compromise on, but this also impacts your budget. You have to take the time to make sure you have security guardrails in place. Maybe you can comment a little bit about what you're seeing on the security front, Carrie. Yeah. It it's interesting. We see that businesses are actually grossly overestimating their security and data protection capabilities. 91%, Lance, 91% of the businesses that we survey believe that they're capable of preventing or mitigating data breaches. But in fact, more than 30% of them have experienced attacks that cost the business in real dollars, whether, you know, it it could be dollars, it could be lost productivity that then costs money. So, clearly, there's a disconnect there in terms of, you know okay. You think you can protect your stuff, but you're not actually able to do that. And so planning for that investment and what's the most, the most streamlined way, the the easiest way to achieve the best possible security? I know that a lot of these, global file systems, like what Nasuni offers, have a lot of embedded features that allow you to eliminate some of your, external or third party protections because it's embedded right into the, solution. And I'm finding that to be a nice cost savings with equal to, if not better protection than some third party providers, can offer. I don't know if you wanna talk a little bit about some of those, protections. Yeah. No. Absolutely. I think you bring up a good point. You know, the the the the the idea of, cyber storage or or or cyber protections and cyber resilience built into the storage layer, eliminates the need for, typical, you know, call it traditional unstructured file, backup and recovery systems. And if you think about what the average company spends on backup and recovery technology and hardware and the software licenses, huge savings. So, that actually represent bring up an interesting point there. It actually brings up that the fact that that is in effect, when you as you rethink your storage model and the delivery of data and AI, you're actually dealing with some technical debt there that you can eliminate. And when you eliminate that those costs, those can go back into funding your AI initiative. Right? You're you're, you can re re repurpose those funds and drive them in a whole better in a better direction and you have better security. And and I think very and faster recovery, which is really the big thing. And let's face it, at the same time, you're offering better protection in many cases. You know, we know that firms with revenues greater than, I believe it's $2,000,000,000, when they are attacked by ransomware and let's face it. It's not an if. It's a when at this point. The attack can easily cost them over $10,000,000 to mitigate. That's crazy. You need to invest in security to do the best that you can in order to prevent these types of things from happening. And even then, we find that it's frequent that attacks are getting through. I mean, we need to do the best we can. Attacks still happen. But the nice thing about the global file system is it gives you that recovery option with all of the immutable snapshots and whatnot to be able to recover that data, fully, if not, or almost fully, if not fully, and be able to restore the business quickly. Do you wanna speak a little more to Yeah. That's, an excellent point. The the the the fact of the matter is the big issue with an attack is when it happens. Right? And it could be ransomware also can be, you know, just a a user doing something, you know, wrong or and deleting whatever. The the whatever the case, the big issue is time to recovery. You have to be able to recover quickly. So I don't I don't care what it is. The the when you do have, this kind of global infrastructure in place like Nasuni offers, you know, in our case, we have customers that can literally have have had issues and they can recover directories, volumes, huge number of files, you know, millions of files in, you know, in under a minute, like, very quickly. It's it's due to the architecture. And that is a game changer when as you compare yourself to if you compare that to, you know, traditional file recovery systems where I have to go mount volumes and move data from point a to point b and, you know, we should be back up in a couple days. I mean, that that's that, you know, that goes out the window when you have this kind of system in place. There are even some recovery providers that are still talking about recovery time objectives in terms of hours, and global file system brings that down to minutes. So it's Yeah. It's clearly a benefit. Yep. Yep. I've I've you see that a lot. Yeah. It won't have you up in several hours. Like, several hours. Forget that. That's lost productivity, and lost productivity means lost revenue, and that's just not acceptable in today's business climate. Exactly. Exactly. Well, let's talk about some common common budget pitfalls, things to avoid. Right? There's we've kind of brought it into four areas, things to watch out for as you start to, or as you move forward with your AI efforts. The first one is sort of a classic, and I do believe it is related to that high percentage of in of companies that have reported initial failures. They've they've probably done one or all or some of these. Right? The first one is the tool first mistake. We brought we, you know, we we bought in early. We got the AI platforms in. We got the tools in. And we we hadn't really we didn't really have our data in order yet. We didn't have all of it. And, so we spent a bunch of money, you know, 60% of our initial budget and, is is what should have been focused on as a as a recommendation. And instead, they just kinda bypass that and start using the tools right away. So you don't get that return on investment, and you don't and you don't see any success initially, which puts a a bit of a stain on your efforts to move forward with AI. You know, the the next one is attempting an enterprise wide deployment without doing the proper pilots. Sort of the big bang approach, you know. And sometimes this is driven, you know, by executive teams that wanna want something happening now, and there's maybe there's there's some pressure to really show progress quickly, which that's great, but you still have to plan, do your pilots, and start out with, you know, some high impact, low complexity examples and use cases where you can show success and show progress and be methodical and have all the right guardrails in place. The other one is the the set it and forget it mentality. Yeah. We did, you know, we we we rolled it out. We were we were, we we got things use it. We've got users using it. We've got, you know, we're seeing some early success. This is great. But then the next shiny thing comes around the corner, right, which it always does, and you've underestimated your ongoing operational cost. Right? So the the answer to that as part of your plan is to, you know, just planned for 40 to 60% of your initial investment annually for maintenance. Right? This this there is going to be ongoing maintenance as we discussed earlier. Finally, the last one is is, related to talent. There is a, you know, organizations are finding themselves competing for, scarce AI talent, and, you know, they're that's just a that's just a factor in in our environment today. So the the opportunity here though as an organization as you're building out this plan is you think about how you're improving your processes and the implementation within your across your infrastructure is to upscale and and build domain expertise, build build that understanding as and it really that's a win win, obviously, for your employees, for your organization to be able to to to kinda grow with the organization as you roll out AI. Yeah. We see that talent shortage crisis a lot of across a lot of different, aspects of IT today. And I'll just put my own little plug in there. In addition to companies upscaling, I hope that there's a few higher ed organizations watching us, watching this, session later because I think that there also needs to be some change in how, education thinks about these technologies and prepare students to enter a workforce that's, you know, driven by AI. I don't think a lot of them are still going about, even technology driven tech tracks, STEM tracks in, in undergraduate work, and it's it's not keeping pace with, an AI enabled world. Right? And so I'd also like to see there be some change in that regard as well, to empower the next generation. You know, we've got a lot of boomers who are starting to retire. We've got, even the Gen x, my own generation is, you know, getting to the age where in the next ten to fifteen years, they're gonna be thinking about exiting the workforce. And what do we have coming up behind it to, to really, replace them? You know, we really need to start to, think about how we're educating students in terms of IT and technical work. Yeah. That that's that's absolutely true. It's a it's an it's an it's an evolution that's taking place and it affects all, all, all folks in the workforce no matter where they're just beginning or if they've been at it for for, for a few years. Yeah. Yeah. No. Good point. Let's let's let's talk about how you can be successful. So we put together a handy little checklist for to make sure that you're gonna be successful. And there it is, that first one right out of the gate, planning. Gotta do it. Right? Many of you would probably say, well, I've done a bunch of planning. Well, just make sure you've got everything planned out. Right? Audit your data estate, especially your unstructured data. Typically, that's 80% ish of your of your total data in your environment. So the vast majority. Audit it. Understand what you have, and then and then quantify, the storage and compute impacts of AI. Right? Where where you want this to be? What what's gonna be the load on the system? What do you need to invest in from an infrastructure perspective? And and we talked earlier about, you know, considering edge locations as part of that for latency sensitive applications or, frankly, latency sensitive users. Right? There is a lot of value as part of this checklist for in establishing a single unified, file structure and having it of your unstructured data in a single location. That, is you know, if you're doing research in this space, you will see that that comes up a lot. That is a that is a foundational element to being successful. And then a lot because you then have your data in a single location, it makes governance, it makes access control, it makes administration much more efficient. Right? And it it really gives you, a a chance to to tighten down your your security, your policies, have zero trust policies in place. And then, as part of that checklist, they, the other one is to then to kind of map out and pilot those use cases. Some of these can happen in parallel. Right? As you're as you're, you know, as you're as you're unifying your structure, you're also thinking about, okay, you know, finance is gonna be able to take advantage of this, marketing is gonna wanna do this use case, and and we're we projected what our benefits and our our goals and our outcomes are going to be, and and we address those. And so we talked about strategic talent allocation, making sure the right folks across the organization understand how to use this and where to drive that value. It's not simply a matter of throwing technology at the problem, but they how to how to really optimize that, and, and then, you know, establishing your cross functional AI budget and having accountability in the organization. This is somewhat related to, you know, some of you are using chargeback or showback models, where you wanna be able to, make sure, you know, if if one organization or business unit is really drawing out a lot of resources, then maybe they need to contribute more, you know, their fair share to that cross functional budget. So that have those discussions at a cross functional level and make sure that you've got your budget in place so that IT is not just taking a huge hit in terms of the budget just because, one or two departments are really going crazy with an AI implementation. Yeah. I agree with that, Lance. I'd like to just take a moment, as we near the close of our session to to take a moment to reflect on the AI maturity journey. Right? We've talked about the fact that we outlined four key dimensions that shape that progression, that planning that you discussed, the strategy and road map articulation. Are you really defining your AI goals clearly, and are they aligned with your enterprise priorities and where you want to go as a business? I think that's really critical in order to ensure that the use of AI is for a purpose and not just for the sake of of saying, yes. We use AI. You want it to be, successful and to be able to deliver return on investments. And in order to do that, you really need to align them with your enterprise priorities. And then the key that we've talked about for much of this session session, data readiness. Does your company have that strategic foundation that it needs for AI? Is your data unified? Is it high quality? Is it accessible for AI use? Do you have the data services in place that you need in order to, ensure that that data is prepared for AI, clean, ready to use. Then the regulatory compliance and policy alignment, are are you prepared to meet what really in this environment is evolving governance and ethical standards, not only for the AI itself, but for the data that's exposed to AI? You've got to be able to ensure that you're compliant with all of these policies because, let's face it, when it comes to things like GDPR or HIPAA compliance, the fines can be really steep, and you don't wanna expose your company to those sorts of, pitfalls. Yeah. Yeah. You know, you you're the the I'll add on to that the fact that is that your strategy is also all always evolving. So you have to have regular checkpoints in place where, you know, the the strategy yesterday may not be the strategy tomorrow, and so we get to make adjustments so you make sure that you're not spending money on yesterday's strategy. Yeah. Exactly. And then you're and you're constantly staying up not only with the technology, but with the evolving needs of the business. Yes. Absolutely. And and then finally, that technology implementation piece. You know, we've talked a lot about your testing and your pilots, and those are critical. But you also have to be prepared and have that planning in place to move beyond your pilots to scalable multifunctional AI deployments. Right? You need to be able to scale that, to an enterprise wide solution that really helps drive value for the business. So these are the dimensions that we see that form the foundation of AI maturity. And I would just, question our listeners. Where's your organization today? And more importantly, what do you need in order to move forward? So those are some questions that, it it's always good to ponder on and to see where you are in in your journey. So let's talk let's talk about sort of key takeaways as we wrap up. Kind of building on what you said around, around what what folks need to focus on. That and, fundamentally, we talk about and we've I've seen this elsewhere as well. Yeah. You need a data strategy for an AI strategy to be successful. Right? What what are your thoughts on that? Absolutely. I mean, it's it's table stakes at this point. You know, there's, there's no way to get away without that. So I think that that that that's definitely critical. Yeah. Definitely. You know, we we really wanna encourage you to as we've said here throughout the session about taking a look at your strategic foundation. You know, data for AI, you know, to be successful, data is fundamentally, that is your foundation. Eliminating data silos, ensuring security and governance, and spending wisely. Choose wisely as it was once said in a famous movie. Right? You know, it's a you know, you don't wanna be those folks where you had a a an AI or a GenAI project fail because you had poor data quality, you didn't have your house in order, and you didn't spend the time to really to build that foundation. Data quality Yeah. We know that 43%, Lance, are are struggling with that today. So it's nearly half, you know, are really struggling to, get that quality data and prepare that for for the AI implementation. So, spending on that data foundation, I think, is a strong investment that can be made, right, in order to, help. It goes a long way to helping your AI deployments be successful. Yeah. And if you can do it in a way that drives cost savings as we talked about, you can actually move, you know, set up your foundation and save money at the same time, that makes a lot of sense. Right? Absolutely. We talked about eliminating backup and recovery infrastructure and moving to lower cost storage models. It's it's just a win win, you know, in terms of your IT budget. So a huge takeaway. That that sort of leads me to, just a a sum summation of of Nasuni and how Nasuni approaches this. Right? This is the plug. Right? The Nasuni's three pillars of AI ready infrastructure really is built on having that global unstructured data access, being able to, you know, completely wipe out data silos and be that single source of truth and provide access at the edge. Right? We intelligently deliver data and being able to incorporate and capture data from from local from, regional or, you know, the disparate sources to be able to gather it for AI's benefit. And then the real, the big piece of it that we didn't go into great the detail on, but it is related to that backup and recovery and the speed of recovery, having resilience. Your infrastructure has to be one that can be back up and running in a minute or two, not hours or days. Otherwise, your entire AI turns into a house of cards. Right? It's not built on a solid foundation. One thing goes wrong, it collapses. That's not the kind of strategy that most organizations want, and that's what we are delivering to customers with Nasuni. It's a it's a architecture just happens to be really ideal for an AI implementation. Yeah. So final thoughts here as we wrap up. I think it's important to recognize that AI's potential is only as strong as the data foundation it's built on. You know, without that unified high quality real time data that we've been discussing, AI models truly are limited. They can't deliver the accuracy or the insights needed to drive meaningful outcomes for your business. And that's one of the biggest barriers that organizations we're working with are facing today. To be able to overcome that, I think that companies really have to take deliberate steps as we talked about with the planning. They need to be breaking down the data silos. They need to, integrate, all of their data sources and establish governance frameworks that ensure trust and transparency, really critical. And building an AI ready data infrastructure, it's it's really not just a technical requirement. It's a strategic imperative if you're going to leverage AI successfully. It's what really enables AI to scale, to adapt, and to deliver impact across the entire organization. Yeah. And that's what we're here for. Right? That's what that's that's why we're all here. We're all looking to do that in our individual businesses. Karen, I I wanna say thank you for joining us today. Great conversation. It's always a pleasure to, to chat with you, and I'm really glad you could join us today. And, yeah. Yeah. It's fun. And and and everyone on the on the watching out there, we we hope that you you have some takeaways from this. Even if you walk away with one thing that you can go back and do or or get or got you to think differently about how you approach, your AI budget, how you approach your spend, and where you put your hard earned dollars, then this was this was well worth it. So thank you very much for joining us, and have a great day.