Video: Exploring AI Solutions: GTM Offers for the UK Market | Duration: 60s | Summary: There are three go-to-market offers available for the UK market, including the "Ask your data" solution for accessing and analyzing data.
Video: Crayon: Optimizing IT Infrastructure and Harnessing the Potential of Data and AI | Duration: 40s | Summary: Crayon is known for optimizing IT infrastructure, reducing costs, and managing risks associated with software licensing and cloud services
Video: Cloud Native Global File Data Platform: Revolutionizing Traditional File Environments | Duration: 31s | Summary: We provide a cloud native global file data platform that helps customers move away from traditional file servers and consolidate their file environment, including backup, disaster recovery, collaboration, and file analytics.
Video: Nasuni: Snapshot Technology for AI Data Resiliency and Recovery | Duration: 94s | Summary: With Nasuni's snapshot technology, data is backed up to immutable object storage, ensuring resilience and recovery for AI-powered business processes
Video: The Value of Dark Data: Uncovering Hidden Opportunities | Duration: 85s | Summary: There is a concern about unstructured data becoming "dark" and inaccessible, but it still holds value for AI strategies.
Video: Accessing Nasuni: A Seamless Interface for Object Storage | Duration: 69s | Summary: Nasuni provides data security, recoverability, and global file acceleration through various features and interfaces, including an s3 edge interface.
Video: Unlocking the Power of Hybrid Cloud Storage: Scalability, Performance, and Cost Efficiency | Duration: 70s | Summary: Our hybrid cloud storage solution combines the scale and economics of the cloud with the performance of on-premises NAS and file shares
Video: Accelerating AI adoption with Crayon's powerful solutions | Duration: 71s | Summary: Our AI services offer a multitude of use cases, prioritizing and implementing solutions to help customers kickstart their AI journey with us.
Video: Accelerate Your AI Strategy with Nasuni File Data Platform | Duration: 71s | Summary: Attend a session on how the Nasuni file data platform can support your AI strategy.
Video: Nasuni and Microsoft Copilot: Empowering Data Management | Duration: 75s | Summary: Nasuni works with Microsoft Copilot and offers capabilities for custom Copilots and integrating with the 0 365 ecosystem.
Video: The Rise of AI Solutions: Trends and Innovations in the UK Market | Duration: 47s | Summary: We're seeing an increasing demand for AI solutions driven by economic trends and the need for innovation and efficiency
Video: Bridging the Gap: Structured and Unstructured Data | Duration: 65s | Summary: The gap between structured and unstructured data has widened, with unstructured data making up a significant portion of the data footprint.
Video: Advanced Security and Data Intelligence:
Securing and Streamlining Cloud-based AI Strategy. | Duration: 90s | Summary: Nasuni offers maximum security capability to protect file data and AI strategies, reducing downtime and time to recovery. No changes to apps or workflows required. Consolidate data for a seamless AI strategy.
Video: Enabling Data Resilience in AI: The Nasuni Advantage | Duration: 94s | Summary: Nasuni ensures data resilience with its snapshot technology, providing backup and immutable storage for AI-driven processes.
Video: Future-Proofing Organizations: AI and File Data Strategies for Success | Duration: 3200s | Summary: Future-Proofing Organizations: AI and File Data Strategies for Success
Transcript for "Future-Proofing Organizations: AI and File Data Strategies for Success":
So hello, everybody. Thank you for, for bearing with us for a couple of minutes just while we, we filled up the room. It looks like we're pretty much well at capacity now, so I hope you're, you're all sitting comfortably. Say hello to the guy next to you, but, yeah, hello. Good morning to those of you joining from team America, and those joining from good afternoon to team Europe. That does make it sound somewhat like the Ryder Cup, but I assure you that you've joined the Nasuni and Crayon joint session around AI and file data strategies for success. I'm sure some of you'd rather be at the gulf, but, either way, you've got me, and either way, thanks very much for giving up some of your time. So to kick things off, I'm Harry Cordigill. I'm an account executive here at Nasuni. So I'm pleased to be joined by Jim Liddle, CIO of Data Intelligence and AI at Nasuni, who'll be leading a q and a session a little bit later on, and also Ian Davidson, AI solutions lead from our partner, Crayon, who will be co presenting with me this afternoon. So today, in terms of content for you, we're gonna be addressing how, 1, the Nasuni file data platform can act as the foundation to that broader AI strategy, and also some of Ian's thoughts and complimentary services for AI success, that Crayon are offering their clients. So I'll be kicking off passing over to Ian, and then Jim will join us, for the open q and a after that session. So please do get thinking for, about some questions for for all 3 of us when we present later. So before we oh, there we go. Before we kick off, I appreciate that some of you will be existing Nasuni customers. So hello to you all again, and thank you for joining us. But, I'm aware that most of you probably won't be, and perhaps this is the first time you've heard from us. So I appreciate we're not a household name yet, so allow me to sort of level set the organization, technology, and the history briefly. You know, why are we here to talk to you today? What relevance do we have to your organization? And, in fact, informing that that AI strategy that that you're here no doubt to help formulate. So how does that platform underpin and enable that AI ready file strategy? So first and foremost, what is the Nasuni file data platform? How do we help? So we're a cloud native global file data platform, essentially. So we help our customers to move away from the traditional limiting in some respects infrastructure of NAS, Windows file servers, and to consolidate not just that primary storage, but also the all the supporting tiers, vendors, infrastructure, software that make up that traditional file environment. So backup, disaster recovery, ransomware protection, collaboration, even file analytics. Right? It's a big stack of infrastructure. So we've been doing this now for 15 or so years at this point. We're not the new kids on the block, and given the limitless scale, right, of of cloud storage, our customers range from 25 terabytes up to 15 odd odd petabytes, so pretty broad range. So let's focus in a little bit more on this. We sit at the top right of what Gartner calls the hybrid cloud storage market. So imagine here if you could get the scale and the economics of cloud, but we couple that with the traditional performance of of NAS and file shares on prem, that land like access. Right? So from a scalability perspective, we start by using the most scalable, durable, and, actually, some of the lowest cost types of cloud storage, and that's object storage for those of you that are that are cloud savvy. So we consolidate all of the customers' environment when it comes to file data into that cheap, deep, highly available, resilient, redundant object storage. So now what's that done for us? That's, you're no longer having multiple silos of of file data to provision, to manage, to patch, to update, to protect, to manage. You name it. It's it's all just held within that gold copy within the cloud. Clearly, this has some repercussions for an AI strategy. We'll come on and and talk about that single source of truth a bit later and what unifying those silos can do for that strategy. Additionally, of course, this frees up IT resource. You know, you you and your teams aren't having to manage all of these different stacks of infrastructure anymore. However, we combine this with the control, security, the local performance, of the edge. We cache that hot data, what's being used regularly by users, day to day in cloud regions closest to your users or in an on prem data center closest to those users. These edge caches are are truly flexible. Run them in as as as as in as your VM. Run them on your choice of hypervisor, reuse some hardware perhaps for that cache. We don't mind. So what that does is, of course, get that data proximity to users for those apps that need it, for that file data that really needs it. Those high performance file workloads, we site that next to users. So, you know, if if if you don't need that local performance, if they're not particularly heavy file workloads, that's fine. Run it all in cloud. Run those edges in cloud too. We have plenty of customers that do that. Finally, around the cloud and on prem, we've got to talk about security. Obviously, it's 2025. So we have maximum security capability to protect that file data as well and, by extension, to protect that AI strategy, those AI workflows and pipelines, that are presumably going to be key to the success of your of your business. So we're reducing the risk of downtime. We're helping you cut that time to recovery as well, quite importantly, if we're thinking about a ransomware event. And we can go into detail perhaps in another session with you. So that's Nasuni. I think I should point out here, we're also talking about no changes to to apps or or user workflows. We're presenting what's that object storage in the cloud. We're presenting that in a consumable way to both apps and users through the protocols that they're used to, through SMB, through SIFs, through NFS. We're converting that object storage into something usable. So there's very little legwork to do from or next to no legwork to do from an application perspective. It's purely a case of migrating that data to cloud, which we, of course, can help you with. So once we've moved all that data to cloud with Nasuni, and hopefully you you choose to do so, we've consolidated that naturally for a data intelligence, and AI strategy. So just to summarize here before I I move on to how this how how does this feed an AI strategy? How does this act as the foundation? I've mentioned public cloud, of course. We're a multi cloud vendor as well. Your choice of of of public cloud hyperscalers, we can give you a common management plane, a common file system across any number of them. You may well want to cite some data within AWS. Maybe there's a particular service in AWS you wanna make use of in a particular region in your distributed organisation. That's fine. Use Nasuni as well across the top of that whilst the rest of your data sits within Azure. We give you that flexibility. Briefly on on the security point, before we move on, I think it's it's it's important to point out, we offer infinite, immutable snapshots for that data. That data is protected down to RPOs of every 5 minutes. We're a versioned file system, and so that has real world benefits for recovery. We're moving pointers around the file system, not performing some of those big restores that you you might be used to with more traditional backup models. So we're able to recover millions of files, in minutes just by repointing you through the file system. And that's Nasuni. So, yes, thank you again to our existing customers for bearing with me there. Now we're all up to speed on the security, consolidation, performance, and limitless scale that the file data platform offers. So how does this then fit into a wider AI file data strategy, and what are the possible architectures? So if we take a look at this this pyramid together, so what what's happened over the past maybe decade or so is that we've seen this split, this gulf widen between structured and unstructured data within the enterprise. So structured data, I think you'd probably agree with me that we've all got a relatively good handle on at at this point, or at least the tooling is there and has been there for some time to help you, to access that data, manipulate it, clean it up, make sure it's in a good state for informing an AI strategy. Hadoop, Apache Spark, Snowflake, unstructured data, however, which is what we're talking about today, perhaps not so much. You know, you'll see in this pyramid that for most, and this is our view, but also that of Gartner's and probably within your own environments, is makes up a a big majority of of of a customer's total data footprint. 70, 80% of the numbers we see quite commonly. So all trends actually moving forwards point to this area being the area of the most growth moving forwards as well, so this problem is set to compound. Managing this effectively, it's a challenge. The sheer volume for one thing, obviously, but also the different types of data, office documents, Adobe Creative content, rich media, CADCAM in an architectural, situation, or team center in in the term in terms of manufacturing. There's no one size fits all here. All of different different workflows, different, sizes of file, different types of file, different requirements for retention, this quickly spirals into into a headache. Digital transformation over the past 5, 10 years has has done wonders for for productivity perhaps. However, when it comes to the IT team and when it comes to informing an AI file strategy, it's caused a bit of a headache here with the volume, breadth, signs of these file types. So now it inevitably ends up in silos. So, you know, on on this slide here, I'm sure there'll be some nodding heads around the room that this is pretty typical for for for a distributed, organization nowadays in terms of what those silos look like, across your different offices, regions, data centers. So that structured data has become more organized that I spoke about earlier. There are workflows to get that into Databricks or or into Azure, and and get working on that data quite quite effectively. Unstructured data has been kind of stuck in these silos for for decades. This probably looks quite similar to, you know, the file environment when you came into your industry, when you came into your job. That's no criticism, of of of your environment. That's that's just the limitations we've been working with with on prem file systems, and the limitations of of that hardware. Right? So there was no other option. So we think the legacy NAS, Windows file servers, and then multiply that out across the regions, whether that's, yeah, regionally or globally. So let's think about the challenges you have in protecting that, scaling it, managing it, backing it up, analyzing these disparate datasets. It doesn't really it doesn't really tell a good story or lay a good foundation, does it, for for informing an AI strategy? Because, look, all of this has value to to your organization. How are we going to consolidate that and feed that into pipelines? AIs, as I've said many times before and you will have all heard, from others, we're looking for a single source of truth. That's what this is about. And if New York's telling a different truth to Tokyo, how does this fit into your strategy? And when we talk to our customers about unstructured data, dark data inevitably, comes up really quickly. Right? Data that's not being accessed. It's not on someone's desk at that time, at that very time. And the the customer data, the metrics point to this happening, becoming dark within 40 to 60 days, you know, your your mileage may vary. The chances of that data then being accessed after it's been been written, pretty slim, within the foreseeable future, at least. Now that's that's one thing from an employee's day to day productivity. But for for for those of you that are looking to try and build an AI strategy, that's not to say that that data doesn't have value to you. Of course, it does. If we're I don't know. Take the example of, you know, looking at some sort of an AI solution that's going to an AI strategy that's gonna help you, help your salespeople respond to bids, tenders, more quickly. Being able for that AI that AM I AI model wants to be able to refer to a bid from 3, 4, 5 years ago that may have specific relevance to, to that person that's trying to write their tender today. It's still valuable. It's just not hot. And so if we think about traditionally how we would have dealt with that data, that old bid, it's probably been, archived now. And and archive is just kind of another word for for silo. Right? It's been pushed off, and we've spoken about what those silos do to your fit for AI. So you see the different types of dark data here. Every one of you sat in this audience will have some skew on this. Now in terms of, you know, archived data, inaccurate data, it will it will differ it will differ between you. Maybe some of you have been through a project recently to clear up duplicates. You, you know, you you tried to refresh your storage array last year, but we're told you couldn't have the budget, and we're told to clean up your your datasets instead. So there'll be differences here in in how this skews, but broadly speaking, you you know, most of you in the room will will have these sort of challenges. And then remember, this is just one silo. We multiply that out across across all the offices, different regions, business units, you name it. So perhaps some of you will be saying that, you know, you've got a pretty good idea of of what data you have. You know, Harry, I feel comfortable where I am. I've been through different assessment processes. I've got ongoing, services to help me clean, curate, manage that data. And that and that's fine, but that's quite rare that you'll have a cohesive understanding across the whole organization. And, you know, thumbs up to to you, if you do. How hard fought was it to, you know, put yourself into that position? How much effort and energy does it take to keep on on top of that that data governance? So, you know, dark data and and these silos are are getting in the way of your, are getting in the way of your AI strategy and probably leaching quite a lot of time, from skilled people within your organization. So how does Nasuni start to address this? What what are what are the answers to the questions that I've raised? This is our value pyramid, we call it. So we start off at the bottom. This is where our initial, assessment stages are, or perhaps where you should be thinking about assessing that data yourselves. Where get get that data into a more fit state, for AI purposes. And this comes down to sort of the foundational strategy of a global unified file system and breaking down those silos. Now there will be some more tactical projects, some some quick wins that some of you will be taking on where, you know, you don't need to go to this level of date detail for, you know, every single AI project in your organization. You know, you you might be here until the next century. But when we're talking about, longer term, wider strategy here, this is where this approach comes in. Get those quick wins done, you know, build some goodwill in the business, show some value, add some productivity to your users. Brilliant. But, yeah, when we're thinking about a wider fit, this is where this comes in. So, of course, we'd welcome the chance to speak through this, with you in more detail at at a later date. Do let us know, when we can talk and understand, I guess, your organization and and what you're hoping to achieve from from building an AI strategy. This is also where we fit quite nice not quite fit quite nicely with our partners, Crayon. And Ian will be able to talk you through, I guess, some of their proficiencies around building services around these various, steps of an AI journey and ultimately helping you deliver some value off the back of that. So final slide from me before I I pass over, to to Ian to tell you a bit more about Crayon. So we've got to the point now where you're at the top of of this value pyramid, right, fit for AI. So we want to get some of these, these workloads, this data, sorry, assessed, consolidated, secured, and get it into some AI systems and pipelines to get some value back. What do the next steps look like? Where you know, where does Nasuni fit within this journey? So it's all in Nasuni now, if you see on the on the left hand side, which is really good choice, by the way, and thank you very much. From here, right, we're feeding it into some sort of a data lake. You know? And this is where it's gonna start combining perhaps with other sources of data. This is where your Microsoft Fabric, for example, comes in. We're gonna collect that data, Maybe look at it again. Maybe give it give it another step of curation before we start feeding your models for for training, fine tuning, and and ultimately deployment. Right? So from here, I hope you can see that Nasuni sits before all of this. We're at the data layer, not not at the compute layer, not at the training, not on the models, which we're purely the data platform that enables these next steps to happen. So perhaps that makes us one of the few vendors that we'll be speaking to at the moment who are quite open in saying we're not an AI vendor, and that that's fine. It's important to realize the full process flow of of of building this strategy. Right? And, hopefully, you see how migrating silos are valuable, definitely, yet previously pretty unwieldy, silos of of file data into a consolidated, secure, globally accessible file data platform such as Nasuni and how that's a key part of this AI planning journey. So the foundation matters in our view. The architecture matters, and then the Suny file system was written specifically for object storage, and we think that matters too. So thank you for listening to me. Really appreciate your time as as I opened up with at the start of the call. We'd really like to get in contact with you and and talk through some of your challenges and priorities for file, and, ultimately, talk through that value pyramid. So you've got my email address up on the slide here. Do feel free to to take that down and reach out to me. I'd love to continue the conversation. Wish you all the best of luck for your plans and projects in 25. Yeah. Let's let's catch up soon. I'm gonna, stop sharing now, and I'll hand over to Ian. I'm on stage now, hopefully. And yeah. Okay. Can everyone hear me, hopefully? Okay. Thanks for that, Harry, and thanks for the introduction. So, a little bit of background on me. I've spent well, I work for Creon. I lead our data and AI division for the UK and Ireland. I've spent most of my career to date in a combination of software and services sales roles. And before joining Creon, some of my more recent roles have had a focus on AI services and legal and governance issues specifically for commercial and corporate law stakeholders. So I joined Creon 2 years ago and obviously a lot can happen in 2 years and the world of AI has been no exception. It's been on an exceptional growth path since that time as have many of us who have been working in that arena as you'd expect. So on to the next slide. First of all, have a look at some of the statistics on this slide. Now I'm not gonna read them all out as part of this presentation. For those who aren't already familiar with Crayon, Crayon primarily is known for helping organizations to optimize their IT infrastructure and reduce costs and manage risks, and those risks are mainly associated with software licensing and cloud services. But we also provide consulting and implementation services as well, and that's really where our data and AI expertise fits in. So from a purely Microsoft perspective, we're an Azure OpenAI global strategic partner, and we've been closely working and collaborating with Microsoft AI product groups since around 2016 on various, you know, data analytics and machine learning projects. We're winners and finalists in the Azure Data and AI Partner Awards, and we've also won AI and ML Partner of the Year previously on a number of occasions. We've also had a number of nominations and awards in other categories outside of data and AI. For example, in this last year, we just won the award for Scale Solutions LSP, which in simple terms means being a skill solutions partner is the partner's really good at helping big organizations to buy and to use Microsoft products effectively. So we're seeing huge growth based on our strategic partnership with Microsoft in in this area alone. Now like most partners, Crayon are vendor agnostic. So around data and AI, we we do also partner with AWS, and we have global credentials with them as well. In fact, we're I think we're one of only a very few, AWS partners to have those credentials. So we also work with Google, although to a lesser extent in the UK and Ireland. So by far, we have the most success with Microsoft in the UK and Ireland, and that's just the, you know, how things are. But we do need to think about, you know, AWS potentially as valid in the eye partner as well. Now our unique value propositions with our partners are based on a couple of key criteria, and the first one is strong relationships and these strategic partnerships with the leading technology vendors and with cloud partners as well. And the second is our global expertise in data and AI space. We're really uniquely positioned as a data and AI partner based on our ability to execute across the globe, 46 countries or so. So not surprisingly, we're seeing more and more customers that are asking for AI solutions, and this is driven by necessity more than anything, of course, as, you know, the various economic trends continue. Only people that are driving innovation and efficiencies, only those who are adaptable are gonna thrive in in this, new world that's been around for a few years. So the current trends that we're seeing apparent in the UK market specifically at present are, of course, generative AI adoption and document intelligence and automation, AI and financial services, of course, which has been around for a long time, growth in the healthcare sector and overall a growing amount of investment in AI, which continues to be strong and you'll have seen some public sector announcements recently around funding that's available for AI projects as well. So, just moving on to the next slide. I guess one of the key questions to ask really is what do our customers actually want? Like, what do Crown customers want when they're talking to us about AI? And as you might expect, that really varies quite a lot and some people or some customers, they want specific detail on technical integrations and options for, you know, what APIs can I hook up with and what other software vendors can I hook up with, etcetera, etcetera? Some people want training and and some don't know even where they need to start. And what our job is primarily, within all of that mix of stuff is really to try and understand which business challenges the customer is trying to solve. So when a customer says to me, you know, can crown do x y zed? Well, really, instead of trying to answer those questions straight away, what I have to do is really turn it around and say, well, why is that a specific challenge to them? Are they trying to cut costs? Are they trying to increase revenue? Are they trying to optimize some process or to improve some specific service and product? Is that internal or external customers? And as well as that, like, what costs are they associated with that? So what ROI do they need to achieve for the service? What budget do they have to address it? And all these other questions. And most importantly, out of all of those, I guess, or one of the most important things is for AI, where is your data and what governance exists around that? What guardrails have you got around the data? And once you've got all these answers, well, then really that's where we can have a conversation, a sensible conversation around AI. Now Crayon specifically, our data and AI practice is a global function, and we've got centers of excellence in Oslo, Vienna, and in Singapore, and that gives us our global coverage covering the Nordics and the the US with Oslo, Western Europe region, and Central Europe with Vienna, and for Asia Pacific with with Singapore and the Middle East as well. So we've got specialized sales teams in all those different areas and I'm one of those people representing the UK and Ireland, as I said at the start. So we focus on the different solutions that are really specifically different industries. For example, for Krayon in the UK, our main focus tends to be our main customers tend to be retail, financial services, and health care, but we do cover, you know, manufacturing, all the other verticals as well as you'd expect. So usually, the way we work is, you know, once we have engaged in a proof of value or proof of concept with a customer, we can really, provide business changing solutions for our customers once we've got a good understanding of their exact business requirements because that's really what it's all down to. Now in our centers of excellence, we're structured in such a way that we can cover all the relevant AI expertise areas. So next to language technology on that slide, we've also got computer vision and decision intelligence or data analytics, if you wanna call it that. And that way we can provide maximum value by extracting all the different types of information from every data type wherever it is in your organization and then provide insights and predictions for the various scenarios that come along with all of that. And the basis for all of these AI and applications are covered by our data platforms team who can provide data to state modernization, management, and pipeline engineering and machine learning operations around that. So, as you can imagine, there's there's quite a good integration there in terms of what Nasuni is doing. Now a lot of our work at present is focused on leveraging large language model technology for what we call ask your data scenarios or content summarization and augmented content generation projects. And we've got, as you can imagine, a multitude of use cases with our customers, and we can work with your internal teams, to help prioritize and implement required solutions. So one of the the big questions, you know, when when I show this slide, people currently come back with and they say, well, that's great. You've got a whole bunch of different teams and but, you know, how do we get started with with Crayon? Like, what offers do you have that would allow customers to get started on their AI journey with Crayon once we've thought about all of these other things? And ultimately, you know, as I said, we've got an endless number of customer use cases, but when we're talking with customers, it's kind of important that we can suggest some initial offerings to to get their attention. Now it may be that their actual requirements turn out to be different. However, the vast majority of cases can be addressed in some way by one of these three offers that we have currently. So, these are the the GTM go to market offers that we have for the UK market, and we do have one pages available for these as well, which we've created. So we can make those available after this call as well. But I'll I'll add a little bit more detail into each of these now, in in the slides that are upcoming, and, hopefully, it just gives you an idea of the type of things that we're doing around AI services and so on. The first one which I've kind of indicated already is called Ask your data and it should be familiar to most people. Sorry, I've got most lines too fast. It should be familiar to most people. Ask Your Data basically means that you've got a question and you've got a bunch of data somewhere in your file storage or elsewhere. And so using an AI application, you can ask a natural language question or query instead of running a SQL query, for example, to get your answers. And it's one of the most common use cases in the market at present. Now in practice, Ask Your Data Gen AI solutions really aim to stream decision making by allowing anyone in the organization to interact with the data as if they're having a conversation with an expert in their field. So if we move to the next slide just a bit more detail around it. Really you you may hear it referred to in conversations as r a g or rag and that just means retrieval augmented generation which in simple terms means pretty much the same as the explanation of Astra data that I've just given you. And it makes it a very useful tool, or useful solution. It's also a little hard to differentiate in the market, but we do have some excellent use cases and one of which I'll talk about in a moment. The next slide that I've got coming up here really talks a little bit around the way things are done currently versus the way things can be done. It's a slightly more technical version of the explanation. I'm not gonna dwell on it too long. It's just a useful slide the customer may wanna see. It shows on the top line the current manual process of doing things compared to, an alternative AI driven method either using a web based chat interface or through a specific business application to ask your data questions. In terms of starter offers that we have around this, how do we get started? Well, we basically offer a simple pilot engagement to customers where we can implement a quick proof of value, proof of concept against specific requirements in their own environment. Now these engagements can attract funding from Microsoft through AWS if the customer's eligible. But if not, it's still a very cost effective way to prove the value out of an AI solution to internal business stakeholders where their needs aren't going to be met purely by a Copilot or, you know, other solution, and where they actually need to think about an actual sort of enterprise wide Gen AI solution for that. One example of that is a customer that we had, where just a second. A customer that we had where they had a large amount of support requests coming in from employees. Basically, what happened is they got rid of a lot of their management team. So a lot of the employees, you know, they went from having, you know, I don't know, 20, 30 people having a manager suddenly, you know, one manager had 2 a half 1000, 3000 people working for them, and, of course, that meant they got limited resources to service, any requests coming in internally, and they were really struggling to respond in a timely manner to some really quite urgent business requests, from both internal and external stakeholders, actually, and things had just become unmanageable. So we built and implemented a solution around Astra Data, which integrated very nicely with MS Teams, and that allowed them to interact with the virtual AI agent and it only escalated to a live team member when the question couldn't be answered by the RAG solution that we put in place. So as you can imagine, it's a very, very useful scenario for most companies and it addresses a real business problem for various customer service scenarios, as I said, both internal and external. That's the first of the starter packs. I'm conscious of time so I'm gonna race on through these. The second one that we have is around document intelligence. Document intelligence really refers to using AI and machine learning to automatically process, understand, and manage documents. It can include extracting data, identifying key information, and organizing documents without any manual effort. So it's all around workflow and replacing manual processes, and it leverages tools such as Azure Cognitive Services and Azure Open AI to automate tasks such as reading and extracting data from both structured and unstructured document repositories, and that's one of the things that that Harry was talking about earlier. And also, you know, within those repositories, you could have invoices, contracts, forms. Also the ability to classify and organize documents within those repositories. And, really, the key thing here is improving workflows by reducing manual data entry and ensuring accuracy, of the results that that are being processed. So again, this offer is really about moving inefficiencies from the workforce and in the case of invoice processing, specifically around, you know, document intelligence invoice processing, it can help to speed up cash flow and reduce payment errors both inbound and outbound, which is obviously, you know, a key criteria for a lot of companies. They wanna make sure that they're getting paid on time. They wanna make sure that they're paying their suppliers on time as well. Some brief technicalities of this one. It's just a starting point. Again, it's a 4 week engagement typically for a pilot engagement, and it would be based around a customer's existing business scenarios or problems in a limited POC. So it gives a starting point to allow us to then go forward and build a a larger architecture assuming we met the the the key criteria for them to succeed in the proof of concept. So I've got some, more information. There's quite a lot of information on this slide, but, again, I'm not gonna read all of the details of the slide because that's not very efficient. The reference case study really goes into so much, but I can summarize it as follows. The customer basically had a large number of data quality issues. This was for a large Australasian bank, and they had a high number of invoices that were causing issues with delayed payments. And that caused them a number of knock on sort of domino effect issues, including revenue loss, customer service complaints, and supply chain issues that that I'd sort of mentioned. And on the next slide, you can see, sort of about halfway down in that slide hopefully, what we managed to do is we managed to automate around 70% of their invoice processing with AI And, of course, that gave them huge time savings versus manual input and manual effort. And we actually calculated, with the customer because they actually couldn't believe it. The ROI, for this solution took 4 months to save start saving them money, which is like, once it was put in production. So that was really phenomenal and a really good use case to talk about with other customers and, for other in this case, it's for Microsoft customers as well. So a really good use case, good case study. And it's similar in some aspects to the Astra data go to market that I talked about previously for customer service scenarios, but it goes far far beyond a simple ask and answer scenario because this provides a high degree of automation and workflow as well. Okay. So the last one that I'm gonna talk about, in terms of our sort of primary offers that we talk to with customers is data platform assessment. And it's not technically a new offer for us or probably for any partner in terms of work we've been doing historically anyway. But we've chosen this one as a focus specifically, as a GTM from the UK because this one opens up lots of different avenues into from an AI perspective into ask your data, invoice processing, but also into other areas of the business such as data security and infrastructure projects as well. So if I was to pick 1 GTM to offer to an existing customer of Crown's for maximum impact, it's always gonna be this one to start with and it lines in quite nicely with other data solution requirements as well. So if I, just quickly go through, there's quite a lot on this slide. It's basically a more technical interpretation of what I've just said to you. A data platform assessment should be almost self explanatory. You see from this slide that we take a highly consultative approach to these assessments. And at the end, we we can provide a customer with a road map and recommendations. Now these engagements aren't typically terribly long engagements. Typically, we'll be talking about, an initial 4 week engagement for these again, and that's mainly to keep costs down and also to align with funding that can be available from partners such as Microsoft or AWS. And the the key point of this is to allow the customer to do discovery of various data and infrastructure issues prior to implementing their AI solution. So, yeah, again, we we look to leverage the various funding programs for these if if the customers are are eligible, which in in many cases they are. We have, again, a good example of a customer that we work with here. This was for a Lithuanian transport company in this case who faced some issues with end of life software and a huge lack of in house expertise as well. So we started out with our data assessment package and quickly used the results of that assessment to create a 3rd plan, which allowed them to then evaluate, the, you know, 2 vendors and build a POC, for the chosen vendor and provide ongoing support. And it's now in production. In this particular case, the customer chose Microsoft Azure, for the solution based on our recommendations of their specific requirements and their specific needs. But, you know, equally, we've done other ones where where AWS has has been the winner in that conversation as well. So that hopefully gives you a kind of an idea of the type of things that that Crayon are able to provide. Those are just 3 go to market offers. We do much much more than that. It's just starting conversation points. So if you do have any questions, you wanna talk to me like in terms of call to action, feel free to get in touch with me. You've got my email address on there, and talk to me about your data and AI options. Or, alternatively, if you want to or in conjunction with, you could discuss your requirements with your the student account manager. And, ultimately, you know, the onus is on you to, book a starter engagement and get started with one of these, engagements and, get going with your AI requirements. So that covers my slides. Thank you very much. I'll hand back. Thanks for that, Ian. Yeah. So now we're we're breaking for q and a as you can you can see, up on the screen there. So, I think Ian, yeah, was good enough to respond to one of you in the interim. What have we got coming in now? One here. Can Nasuni work with Microsoft Copilot? Are you, there waiting in the in the wings as well, Jim? If you could walk through the metaphorical curtain onto said metaphorical stage. Yes, that is the short answer. But, Yeah. I'm here. So the the answer is yes. Nasuni can work with Microsoft Copilot. Microsoft Copilot, I guess, is, an analogy these days for, a suite of products because Copilot covers more than just one thing. But, if we break it down into, let's say, Copilot Studio from Azure AI Foundry and, o 365 Copilot, yes, Microsoft can be used with both. The difference obviously between the both is, Copilot Studio is really more about custom Copilots, and, you know, building them the solution yourself with, with Copilot Studio and Azure AI Foundry. And, in terms of, o 365 Copilot, we use a graph connector to allow the data stored in the SUNY to be pulled into the 0 365 ecosystem through the semantic index, with full ACLs and permissions. So that when you then interact through the 0 365 app ecosystem with those, with the data, it will, you know, basically look at your email data if you're using 0 365 email. It'll look at SharePoint. It'll look at OneDrive, but it will also look at Nasuni data when it's forming the answer. And it will only use the dataset that you have permissions to on Nasuni based on who you are. Excellent. Thank you for that, Jim. Hopefully, that answered your question. Certainly quite a lot of detail there, so thanks, Jim. Another one from, Richard Jones. How does Krayon integrate Nasuni storage into an AI large language model? That's an interesting question, and I'm not sure I can answer it. How long have we got it? I think there isn't a a single straightforward answer to that. I think it, ultimately, the, LLMs are gonna be looking for, you know, where our data repositories are held. So, kind of the assumption would be that the Nasuni stuff is is put in place, and we're just pointing at that repository to, run through the the various, data that's being held in that repository. So I don't think it's anything over complex to to answer that question actually, but, yeah. Maybe we need if we need to go into further, like, deep dive stuff into that, we can we can certainly set up a separate conversation around that. Yeah. Yeah. And I I can probably add to that. Having worked with both cloud based LLMs and local LLMs, local LLMs, if people are using Nasuni data, it tends to be the file interface because that's the easiest way to, to get access to the data if you're using LAMA or you're using an abstraction layer like Ollama. Cloud based LLMs, we actually offer, an s three interface through the Nasuni Edge into Nasuni which can act as a gateway to get access to the Nasuni data and plug it straight into some of those, AI solutions. So it's it's fairly easy to get access to the Nasuni repository from a both a cloud based and a local based LLM perspective. Good. Have we got any any more coming in? Just as the time, it takes around to 15 minutes left, so probably time for a few more. How does Nasuni enable data resilience when AI is being used? So more of a, yeah, question around, yeah, resilience. Yeah. So, actually, resilience is intrinsically built into the Nasuni architecture. So one of the benefits of new using Nasuni is, it's snapshot technology. So anybody who's using Nasuni will understand that. And that snapshot technology basically takes snapshots of data and, backs up those that, that data, to object storage, which is immutable. And, you know, if you start to think about where, AI is going in the agentic architectures and agents, you know, ultimately, you'll have, data powering business processes, which are exposed as autonomous agents in an agenda architecture. And all of a sudden, the ability to have, dead resilience at the AI layer becomes really important. As much about recovery as it is about resiliency. And one of the things that Nasuni is really good at is first of all, it can limit the the blast radius because it has some, you know, ransomware protection, elements to it. But not only that, you know, based on the precept that at some point, anybody could get, you know, kind of a threat actor lock out their data With Nasuni, because of those snapshots and the immutability where it's stolen cloud, actually, getting the data back is as simply is as simple as moving the pointer back to a previous object. And that means you can restore millions of files in in minutes, not in weeks or months to get your kind of business processes exposed in this agenda architecture or exposed through an LLM back online, you know, really quickly. So Nasuni actually has AI data resiliency built into its core. Thanks for that. Yeah. An an integral part of the file system and and one of the goals from the very day that it it that it was written, really, wasn't it, Jim? So it's only become ever more more relevant, really. So we we we put the dev work into the right place, luckily, back in the day. Got a question from Jamil Hussain here. So thank you, Jamil. Can we protect a d l s v 2 against ransomware and corruption? Totally answer that one, Yeah. I don't know if they'll to to our security team and just, and just follow the Yep. So thanks for the question, Jamil. What would you describe as Nasuni's drawbacks? That that's that's an interesting question, isn't it? I I guess it could be a a perceived drawback per se in that, obviously, separating that data from from users. You know, if if the work hasn't been done in the first instance to to make sure that those caches are sized correctly, right, on prem, then you've potentially got an issue there of of users relying too heavily on on pulling that data down from cloud rather than accessing it locally from a cache. But, I mean, that's a function of of our engagement, really, and, you know, we're we're pretty adept at at making sure that cache is sized correctly in in the first instance. We typically get 99% hit rate on our caches. So and that being said, you can easily expand them once they are deployed. So that's just that's just one that that comes to mind. But, yeah, don't know if, Jim, you've got any, any other drawbacks of Nasuni, or we'll move on. No. I think the the one that people talk about is, obviously, Nasuni works inside of an encrypted container. The reason it does that is to provide some of those ilities I've just talked about. The the ability to, to make sure that your data is secure and recoverable because of the way that we, we chunk the data up and protect it. And and not only that, the otherilities that we provide around, global file acceleration between edge caches, the global locking service that we have, like Snapshot, you know, utility itself. But what that means is is that when you access Nasuni, in the past, you would you would typically do that through a file interface even though it's stored on object storage. And and one of the things that, you know, we've done to, to make that easier for customers to access, You know, we've we've implemented our, s 3 edge interface so that it can look just like a compatible s 3 bucket. And, you know, later on, later on this year, we'll be doing the same thing for Azure Blob. So even though that might be kinda, you know, deemed as as as perhaps a limitation in the past, even though we do all of those great things is to the reason why people, you know, go away and procure in the Zoom, we'll also be allowing you to access it just as if it was native object. Thanks for that, Jim. Jamil, you're active. So thanks for the thanks for the contributions again this afternoon. Do you support prioritizing restores based on labels slash classification done by, Purview, Microsoft Purview? Well, we do integrate with, you know, Microsoft Purview. That's for sure because we've we've done some work with customers around that. In terms of, how we can do the the the restores, the restores can certainly be done, against, against different file sets. Whether that plugs directly into, Purview labels and classifications, To be honest, I don't know. It's something we'd need to go away and check. Yep. No problem. We'll we'll add that to your other question, Jamil, for follow-up. Any more coming in? I think, Jamil, you were the the last one. I don't think we've missed anyone either. So, yeah, speak now or forever. Hold your peace, and, we'll we'll look to wrap up with next actions. Yeah. Well, you've got our details now anyway. I noticed the instructor's, email address into the chat. So let you know where we are. Please do darken our our door. Give us a challenge. We look forward to speaking to you around your specific challenges and and strategies around around AI. Bobby Clapp, actually, just noticed, joined late. Will we get a recording sent to us? I believe this one, will be recorded, and you can access it, through the link afterwards. But, yes, there you go. Thanks, Annika. Any closing words, from yourself, Ian, Jim? Yeah. Just give us a call if you've got AI requirements around, any workshops. So, you know, I mean, a a lot of customers may be looking at Nasuni, for example, might be just at the start of their journey thinking we're gonna consolidate our data, and then we've got an AI project coming up, for example, or, you know, compile implementation. So if you do get, unstuck in any of that, yeah, just feel free to get in touch. My details are in the chat and on the presentation, so just feel free. Thanks. Nothing for me. I think you both did a great job, so well done. Great. Well, all that's left for me to do is to thank you all for joining and wish you a pleasant evening slash day slash wherever you are in the world. So, yeah, thanks again, and look forward to speaking to you soon. Thanks, all. Thanks, all. Bye.