Alrighty. Let's get going. We've got a lot of folks that have registered to join. So expecting some other folks to continue to trickle in, but that doesn't mean we can't get started. I'm just going to modify my little setup here to help me better do these things. Just give me one little second. Sure. My audio is all set up. Perfect. We can get going. Okay. Great. Welcome to, the future of analytics, the journey to BI three point o. We'll discuss what that means, how we got here, all sorts of fun things in the next fifty to sixty minutes. But by way of introduction, let me start by saying who I am. I am Robert Curtis. I'm the managing director for Interworks looking after Asia Pacific. I'm based in Melbourne. I've, been with Interworks going on somewhere around twenty years. So I have been with Interworks for a very long time. I've gotten to help a lot of different customers of all shape and variety, do a lot of different things with their data. You are joining as part of a series of webinars that we launched with the conclusion of the Snowflake World Tour. So some of you may have been in Sydney, may have seen us at our booth and registered. Some of you might be friends that are just joining. We've got quite a bit of content for you over the next four weeks or several weeks. First off, we've got a white paper that we have written that is available for your download now. The preeminence of the semantic layer for context of why the semantic layer is now the most critical piece of your data architecture. Whereas before it was really more your data analyst in your dashboards. Because you have AI consumers, it changes the entire dynamic. We are doing right now the future of analytics, the journey to be a three point zero that is today. Next week, we're doing the preeminence of the semantic layers. Again, we're talking again about the importance of that. Week, maybe two weeks after that, we're going to talk about AI at scale. So think about this as your guide to building an AI center of excellence, looking across skill building, governance, change management, all of the critical things that often don't get really properly planned out that you need to do to be successful. And then we have a completely new topic called the five essential AI use cases everyone should do. And in fact, I think we have another one in there on September second, which is five essential things you should be doing with Snowflake today. Something like that. I'll update the slide the next time around. Those are all available for registration on interworks dot com. If you haven't and you wanna see these, register. Even if you're not able to attend on the day that we do this, we record everything. We'll let you know when the recording is up so you can watch at your leisure or share with friends if you like. A little bit more on Interworks. We basically do three things. We do data strategy, solutions, and support. What does that mean? Well, if we break it down a little bit further, we help you build the direction. We help you make the decisions. We help you plan plot and roadmap for what you want to do with your data. Then what we do is we help in terms of building the foundations that might be your platforms, your cloud governance policies, the implementation of governments into systems and applications, building the data itself. So the data pipelines, your data platform, the architecture, the modeling, all of that stuff. Once we have the foundations in place, then we can start really adding value. And that is doing analytics or advanced analytics or reporting, building AI solutions, doing data science, but really building intelligence and usefulness into the insights and ideas. And now with AI actions that we can do for our business users. Once we have that done, then it's time to sustain and grow, which means it could be sustaining the applications or tools that are pipelines even. We help we offer managed services to help you with all of those. We can also help support your individual users in building their skills, whether it's in data engineering or analytics or AI or governance or whatever it might be, as well as communities, meaning community of practice, centers of excellence. We can help build all of those things. So however, and wherever you need us, so long as it's when that sort of data life cycle, we certainly would love to help. In terms of a little bit more contextualization before we get into this topic, Interworks is a little bit of an outlier. We're a services integrator or a consultancy. We're not one of the big four. So most of our competitors get big and get sold within about ten to fifteen years. We have thirty years of experience because we really love what we do. And every time I see this slide, I tell myself I'm going to update that twenty seven to be thirty, and I'm going to do it even though I'm going to forget to do it for the next slide that I do in the next deck. So here we are at thirty years. Seventy five, probably I think it's seventy seven of the Fortune one hundred are InteWorks customers, the biggest most complex, datasets, use cases that you could imagine. Some of these are on our, website if you would like to learn more about our case studies, companies like Google, Netflix, Facebook, all the way down to the not for profits that we service here in Australia. We're not looking for a type of customer. We're looking for a type of person, Meaning, let's partner together. Let's do something cool. Let's build long term success. We're not specifically vertical specific. We really like working with great people. I would recommend that you check out our interworks dot com blog. We get probably closer to four million paid views now in terms of just people learning about AI or infrastructure or governance or analytics or data. We've got four thousand plus customers across every vertical you can imagine. Here in Australia, we do a lot of work in mining, in energy, in financial services, public sector, not for profits, health care. We do a lot with a lot of different folks. And of the things that we're the most proud of is we were given this award called the Forbes Small Giants. They picked, what was it, twenty five different organizations across different services or verticals. This is a small company that punches well above their weight. We were delighted to be chose one of those for data services, which is really our excellence, our commitment to excellence in being the very best consultancy that you've ever had. So for all the people that are joining us today, welcome aboard. For all the new friends, for all of our old friends that are returning, welcome back. Great to see you again. So let's get going. I mentioned today that we're gonna start on, we're gonna talk about BI three point zero. We're gonna define what that is and and how we got there. So let's start from the very beginning. The journey until now. Old man, middle aged man, which was what Robert's probably closer to. We'll say young man and then child. So I'll tie all of this in in terms of what this means, but you might be asking yourself, wait a second. You said BI three point o, but you've got four different people on there, the different stages of life. Why four and not three? Well, it's trivia time. Everybody loves trivia. I'm gonna ask I'm gonna show you two pictures. I'm gonna ask you what these two pictures have in common. And if you want, you can throw your answers into the chat. We'll see if anybody gets this right. The first one is not so much this, archaic, desktop terminal workstation. It's really the application on it. So take a look at that little spreadsheet on the application and then take a look at this other picture of this young boy wearing a cowboy hat in a vintage little outfit riding a rocking horse. What do those two things have in common? Any guesses? We don't have anyone throwing it into the chat. Maybe I've stumped you. Vintage from Anna. That is correct. Both okay. Now I've got people that know me that are making a little joke. Thanks, Kathy. Both old. Yes. I feel old every day. The answer is that is me as a little boy back in Oklahoma, I think at my grandfather's house. The answer is both of these items, both the spreadsheet application there and young Robert Curtis are from the seventies. So maybe I am the old man instead of the middle aged man. The reason I bring that up is VisiCalc, which was released in nineteen seventy nine, is really the earliest iteration of business intelligence. And what I mean by that is business users being able to use data to make decision making. VisiCalc, visual calculations, if you like, was really one of those first, quote unquote, killer apps that really justified personal computers in every home and in every business. It's kind of funny because even just a couple years ago, people were like, well, AI is powerful, but what's the application of AI? Well, I'm old enough to remember people saying, computers are great. They could do a lot of things, but what's the thing that you have to have that they can offer you? And VisiCalc among other things was one of the first few things that like, yes, this is a clear value ROI addition to the capability that this PC offers. Super important, super influential. So this is BI zero point zero. So this is pre BI. And you see VisiCalc, the earliest version of Microsoft Excel, one of the longest, most useful BI tools ever invented, Lotus one, two, three, all of those things. And what it did was very, very simple. It did some basic data storage, nothing anywhere close to what we can do today. It did processing. So if you wanted to ask some questions or get some numbers aggregated, you could. It was the first electronic spreadsheet. It could do some automatic calculations like totals and sums. It had very limited ability to do some modeling and forecasting. If this, then maybe this in the future. It was accessible to the standard business user, probably somebody that was fairly progressive with technology versus anyone that could just open it up. And it was customizable with very simple formulas. Now, of course, this was groundbreaking, but in the retrospect that we have from twenty twenty six, there was quite a few limitations as opposed to what we're used to today. Performance was very limited. Visualizations, well, you saw the picture of it. A green chromatic, is chromatic the right word? A green screen that basically had no visualizations, it was a spreadsheet. Lacking in advanced analytics, lacking in real customizations, almost no data preparation whatsoever, and kind of really, you see what you get, but it was a massive step forward. Now, as we go through the different versions of BI, I'm also going to share with you an overlap of BI operating models. I'm going to show you these sort of in order, but that doesn't mean that these operating models don't still exist today. But generally when you had VisiCalc, you didn't really have much of a choice than to be fully decentralized. Someone had it on their laptop or not, they didn't have a laptop, somebody had it on their PC and they would do analyses for themselves. You still see that today. There are organizations that are still getting their head around what they're going to do. And so decentralized BI, another way of saying that is the wild, wild west. You just have a random scattering of people that are data progressive and they just start doing stuff. It is the starting blocks for most organizations that are thinking about data. And because it's fully centralized, you have that pendulum of things that swing to your benefit one way, but at the same time brings a lot of disadvantages and it swings the other way. You generally try to trade them. So the disadvantages of decentralized BI is, well, there's no strategy, there's no governance, there's no cost controls or efficiencies. The centralized effort of building data assets is often non existent. The economies of scale, that kind of thing. But what's really, really good about this is that you get a tremendous amount of agility. You don't have controls, You don't have governance. You can just start doing stuff. And sometimes that's useful depending on the frame that you're in. If you're a digital startup, oftentimes doing stuff and adding value is more important than slowing down and controlling and making sure we're documenting and being safe and secure. Again, it's business by business. But your individual users are the evangelists within this tool, within this BI, within this data model, this data approach. And oftentimes they are the innovation agents. So those people, those early adopters, we still want to empower them regardless of what type of BI operating model, which is sometimes we wanted to treat them differently versus just turn them off the leash and let them run wild like lone wolves, but that's decentralized. How does that look? Well, that means you've got IT doing IT things. And in the old days, they were busy doing other things. And data analysis often fell down to, self-service data analysis, at least often fell down to just enterprising data progressive individuals scattered throughout the business. And again, no centralization, tools were random, no concerted effort to build monumental sort of artifacts like data products. That was BI two point zero and then how that intersects with the Wild Wild West. Now we get to actual first generation BI, BI one point zero. And that's where you get BusinessObjects, Cognos, MicroStrategy. And really, that was one particular thing that really kick started this. And that was the arrival of the very first ever data warehouse, Teradata. I can assure you Teradata, maybe not alive and well, is definitely alive and still kicking. Even just in the last month, was working with a customer that was looking to get off of Teradata and Teradata was still doing things. But because there was a centralization, a repository with advanced powerful features, that meant that BI could then hook onto a data platform that allowed it to do a lot more things. For instance, slicing and dicing, and processing your analytics. Scheduled, predefined, prebuilt reports that could arrive to you. Business friendly abstraction layer, I. E. The semantic layer. So all of the stuff that your DBAs are doing, it doesn't have to appear that way to your business users. It could have more friendly terminology, more friendly names, more friendly definitions. Overwhelmingly maintained and governed by IT, that probably hasn't changed too much. Although IT might be now quote unquote a subset of IT professionals such as the data team. Information technology sort of consumed everything here. And then obviously with greater capacity came greater complexity that you could do with your data and therefore your analytics. Now limitations, lack of agility and speed, obviously. And that's still something that even in BI two point zero, we're still going to struggle with. Long queues, meaning, a lot of people need stuff and only IT people can build it. It means we're waiting on them. Only batch data, only structured data, unstructured data is completely invisible, high costs, but again, you don't have much of a choice. It's the only game in town and high barriers to get in and make this useful. Now, we overlay this in terms of what the BI operating model might look like, this would look very much like a centralized BI setup, highly towards the middle. Another way of thinking of this is the report factory. And just like we saw in the Wild Wild West, the centralized operating model is exactly the opposite. So all of the benefits that you got from the decentralization you lose, but all the disadvantages you address. So with BI one point zero or the report factory, you generally saw very strong alignment on strategy because IT was setting it. Now they might have their CFO or their CIO. Today, we have a lot more exciting terms to be more specific in terms of the different people that touch data. But those people were saying, this is what we're doing, and this is what and this is why. And the CFO who often would lead IT would say, and I have the money to to put the mandate on it. The other benefit of being highly centralized, you could be highly governed because you don't have people touching your data that aren't fully trained, developers, engineers, software folks, whatever. The challenge is you gave up almost all of your agility because you cannot have a full capacity to meet everybody's demands as they might have it. You have to put it into a queue and tickets, and people that are more important get their stuff faster. So the development time, the pace of insight generation slows down dramatically, as well as the development of data products. All of these things slow. IT by definition is separate from the business. They are a cost center, they're a production facility, they are the technology arbiters, they are the platform owners. And so as a result, oftentimes they don't understand the analysis and reports they're building. It has to be transcribed, translated to them, and that could take several iterations to get it exactly right. So ideation and innovation, the result was quite low. What does this look like in practice? Well, all of the distributed people that were data workers are now embedded into IT or quote unquote data team. You can see there's these cues of data products that have to be built that then drive dashboards and reports. Everyone in each of the business units is looking up, can I have my report? Inevitably, you're going to have business users that have those frustrations because back in nineteen seventy nine, I could start to self-service my own data needs. They're going to say, I'd like to move faster, Which gets us to BI two point zero. This is where Tableau came and kind of revolutionized everything. And this whole idea of data for the people. You do not have to be an IT person to work with data. You do not have to be a statistician to do analysis. Other big breakthroughs that came with Tableau near real time data near is obviously the operational part of that definition could mean a lot of different things. Every fifteen minutes, every fifteen seconds, You can start approach advanced use cases, even though Tableau had some limited statistics and forecasting, could integrate with R or other tools and really do some more advanced stuff. The big thing was just how interactive and visual these things could be. And just how easy it would be for somebody that was, let's say Excel savvy to get in and start doing stuff with Tableau or Power BI or Qlik or whatever. Tableau was the first, and they each kind of have a little bit of a different theme to them. Power BI, for instance, to this day is still a bit more of a developer's tool, whereas Tableau is more of a business user's tool for better or for worse. There was, the ability to integrate better with mobile, mobile phones, mobile devices, laptops, that kind of thing, embedded applications so there was more agility in how you used it. And there was just the first twinkling of AI. Explain this, Mark. Ask data, those types of things. Very basic, very limited, sometimes not helpful at all, that those were just starting to surface up. Now the challenge with BI two point o, which is important to understand why we're going to put BI three point o and why BI three point o is taking this particular shape, is that tools like Tableau and Power BI, etcetera, made it very easy for business users to create dashboards. So much so that they quickly outstripped the ability of data engineers in IT to build good data sources. So every time we wanted to build a dashboard, we probably ended up doing some data preparation for the dashboard as well. And every time we had a new question, well, we didn't have good data, we didn't have application, we didn't have other ways to interrogate this data other than dashboards. So, baboom, what may have been a hundred dashboards in one point zero became ten thousand dashboards or even a hundred thousand dashboards in some of the clients that I've seen. Dashboards for everyone, dashboards for everything. That creates a substantial amount of work. We'll talk more about that in a second. Then you can see with these dashboards and the way that these dashboards were constructed, the logic started to creep more and more and more into the analytics layer. So into Tableau Server, published data sources, Tableau Extracts, into the workbook and embedded Tableau extract or in the worksheet itself. This is the calculation for how I determine revenue in the workbook. That works if human beings are your primary mode of consumption. We're writing these reports so that human beings can see them and make decisions. It doesn't work when other types of consumers are there. In fact, it breaks quite quickly. Now, even though BI two point zero meant people that were savvy could go build themselves dashboards, it still meant there were queues and people waiting to get their insights. So let's overlay this onto federated or self-service, meaning everybody can build stuff. They build the dashboards and IT tries to keep up with building data products. But inevitably, your federated or your spokes in this hub and spoke model outstrip the ability to build data products or the agility they need, they just start building off and you start to get these long tails of siloed business logic. There's some advantages here. So because the center is involved, strategy can be sourced from the spokes, but then centralized in terms of what we're doing. Governance can be delegated, but centralized in the tooling. A lot of times these federated things is we'll choose the platform, we'll build the data, and then you guys find the business value in it. But we're gonna give you the rules of the game, meaning governance, access to data, privacy, security, usage policies, those types of things, you have to abide by them. We'll do everything we can to automate those, but there's still going to be human beings having to make human being decisions of common sense. The challenge that you have is, and you can see this with everything being purple, is that everything kind of ends up everywhere. There's stuff that's in the center and there's going to be stuff in the federated teams. You just have to figure out is that good, I. E. Agility, or is it bad, meaning governance? And in the future for BI three point zero, where does our business logic actually sit? So if we look at about this in practice, we've got data workers in IT. Their job now is just to build datasets, build me data products, give me gold level products, platinum level products. And then in the business units, not everybody's gonna have the ability to go and build a Power BI report or a Tableau report. They certainly can use them, but we still have to have people that are knowledge workers, are data specialists that can build analytics. So they still have queues, but the queues are now broken out by business unit or team or function. And IT people or other Tableau people like me would build these reports for other people and then some for myself. But because I'm embedded in the business, my ability to understand what you're asking is going to be better than if an IT person built it. And it means I can give you a bit more customization and we can get to a better result faster, but still you're waiting on me as your associated data worker to support you. So if we look at this in summary, we've got pre BI foundations, which is your ViziCalc, starting from the 1950s into say nineteen seventy nine and the early stuff, early eighties. Foundational ideas, it really led to the explosion of the PC, which then they led to the explosion of processing power, which then of course obviously led to information networks, which led to the internet, which led to cloud. All of those things start here. No data quality, limited logic, isolated systems, but stake in the ground had been said, wow, we can do stuff, cool. We then get to BI one point zero from the 1990s to the 2000s. Now we can do more aggregations. We can present, can ask questions like, when did this happen? What happened back then? Generally IT driven, generally centralized. And you had the first data warehouse to start to appear, which really unlocked more features and functionality. So what happened when it happened? So now the question is, I have lots of information. I know what happened. I know when it happened. I've got some dimensions on slicing and dicing it down to different aspects, but what do I do about it? What decision do I make? That was often a very human intuitive analysis that human, okay, this is what these numbers say. What do they mean? Then we get to BI two point zero, which is from say twenty fourteen or twenty ten, somewhere in there to really starting to wind up to about right now. And the questions are, well, how did this happen? Why did this happen? You start to get deeper analysis. It's driven by the business. Better visualizations, better storytelling, self-service, better answers. So now the question becomes, how do I get better answers faster? Because it still requires human beings to prepare this. And a lot of human beings are doing a lot of work to overcome the shortages of your data that's in your data warehouse to then build this value chain through the analytics layer to get those answers to where they're automated. That doesn't mean they're consolidated, it means they're automated. So we have to ask then, based off of this trend that we're seeing, what is next and why? Well, this is where AI comes in. So AI needed a couple of really important things to happen before AI could become a thing. One, obviously we needed models. We need vector databases and large language models and all of the stuff that's really been groundbreaking revolution over the last three years. We needed cloud computing. So when Amazon started building this super warehouse of stuff, you could order that all this extra compute and like, we should sell this compute to people because we only use it at peak times. There's all these other times we're not using it. We may as well monetize it. So people were able to start buying compute and processing and and computer usage from the Internet. That's basically what it was. And as a result, the elasticity, the cost effectiveness of growing really, really big and then shutting it back down became achievable to do really cool stuff without having to buy forty five server farms into your backyard, which then gave us the ability to start looking at other things. So conversational analytics, the ability to talk to data, natural language using vector databases can be converted or generate new language, new conversations. All of these things are groundbreaking. Proactive and contextual data, augmented workflows. These are all things that are starting to add and come together to really create a foundational revolutionary moment in the way we think about data and how we use it. Which brings us to BI three point zero. So here we have AI holding a little platter of questions and answers for you. AI is the data worker that was embedded your department. But rather than being in your department saying, hey, Bob is the great person that comes with analytics, just go over to that person's desk and see if he can help you with your report. Everybody has an AI agent. It's integrated directly into the cloud data warehouses. It is directly integrated into your everyday workflows. It gives you more ways of working, more options to ask questions, more places to ask questions. And if data is really, really good, which is the big caveat here, you can ask a lot of questions. We use Claude, and we have a lot of data at Interworks, and we started integrating it. So now I can just go to Claw to start asking questions about our sales pipeline, our clients, our delivery, our projects. I don't need to go into analytics tools anymore. I mean, have reports, but there's about three that I use as opposed to the fifty that I would routinely use just for myself back in the day. You have this idea of agentic AI, which means you can give a task to an AI bot and it'll go do it. That is a very different experience than a dashboard, which is just there to tell you something. As a side note, if you're worried about governing a hundred dashboards, imagine trying to govern fifteen hundred agents. And a lot of these pop up and spring up over the organization, just like dashboards do unexpected unplanned by third party tools, your quote unquote shadow AI by developers that shouldn't be doing this, or you didn't know we're doing it. And they can actually change stuff. So it becomes certainly something that you want to sort of get your mind around. So lower barriers to insight generation. I don't need to understand Tableau. I don't need to know DAX and Power BI. I can just go ask questions in natural language. And AI, again, is getting a lot better at giving me answers. We can still support it with a great semantic layer and great context layer. And those things all come together to give us the best result. And then analytics as an application interface, longer is it necessary for analytics to be here's your answers. And then you swivel to a different monitor and start doing stuff. You can now through Agintiq AI, build actions directly into your dashboards and reports. Do this, click this button. This is the idea of three point zero, the next evolution of AI augmented analytics. So the limitation here is AI is only going to be good as your data. So curated datasets, the data engineering part of your in of your journey becomes absolutely mission critical. It's no longer the dashboards. It's now your semantic layer. And that's one of the topics we cover in a upcoming webinar. So here's the full view. You can see pre BI, wave one, wave two, wave three, BI one point zero, two point zero, three point zero. Now this is what we're looking at. We are here in BI three point zero. I don't know if I would have said this even six months ago. It was very much the promise is there, the foundations are there, but you can do this today, which means I can anticipate your questions and I can enrich them. I can facilitate the what, when, how, why, to what extent, what next, all of those types of questions can be asked. AI is my personal data worker that I can then send them off to go do anything, and you have yours. And Susie and Larry over there, they have theirs. It is highly personalized to your persona or to your cohort in terms of the types of answers you're getting. So if I'm an executive, I could say, why did this happen? Aggregated high level directional result. If I'm a product owner, why did this happen? Detailed, specific. Again, AI could build you that bespoke experience. Better questions, faster answers. So the question now becomes, how do I solve data readiness at scale? Because I need curated data sets to really get the most out of AI. That becomes the critical thing now. Just while we're looking at it, so this was what we looked at while we were talking about BI two point zero, federated BI or self-service. This is what it looked like. So we had queues up in the data side. We had queues down in individual business units and distributed knowledge workers throughout. When we think about this from an perspective, everybody gets an AI worker, including the engineers. There are already significant breakthroughs, like true crazy breakthroughs from software engineering. And software engineering is what you would call like a leading indicator in terms of what's going to happen for data engineering. They're getting ten to one, thirty to one accelerators on software engineering. That is coming for, that is going to be here for data engineering. Right now, if you think about this in terms of layers, bronze, which is staging data, getting it through APIs, that's probably the easiest thing for AI to figure out, particularly when it is something that can be documented and contextualized. AI goes, ah, you want to get data out of Salesforce? We'll document. I can go get that for you. Workday payment gateways, whatever it is, much faster. Silver, a little bit harder, gold, a little bit harder, but those efficiencies are coming. I've done one report, seen some other stuff that's looking like in the next, say three to four or five years, gold and silver will be somewhere in the neighborhood of sixty to seventy to maybe eighty percent accelerated based off of AI, helping data engineers do their job faster. And likely it will be less data engineers and more really, really smart data architects, senior data engineers that understand how everything fits together and works. And they've got the AI going and doing the minutiae, the coding, the tasks, the little stuff while they orchestrate, which is exactly what you're seeing in software. Down in the business level where they're doing the analytics, again, every single person has AI at their command. So I could ask it questions. I can get it to build apps. I can get it to go and investigate, do agentic stuff, take actions for me. I'm accelerated by it. And there are no more queues. It doesn't mean there's no more dashboards. It means that for my organization, let's say a thousand people, maybe I have a hundred dashboards total. Well thought out operational reports, executive summaries, beautifully thought out told stories that AI wouldn't be very good at trying to come up with on the fly, but just answers. What happened with this account? Who is my main stakeholder in this blah, blah, blah, blah, blah? AI could just go and grab you the answers. Literally, it's just having a conversation in Slack. So what does this look like? So if we were to take the old workflow of building a data product all the way down to data insight, we start with our data, we build a data pipeline. Again, that's manual. We package it up. Once we build the logic, the packaging of it, that can be automated. So you get a little gear there. A data engineer would do this. Then somebody needs to build a visualization or a dashboard because again, in the old way of working, this was how we consumed everything as our insights. This was the common currency. An analyst would do this and then a user would consume it and have an idea and insight or take an action. But the problem is, is what if they had another question or this dashboard didn't quite answer what they needed? Well, we'd have to go back to the data and potentially re orchestrate one or all of this entire process. But what if the data didn't exist in that data, that data warehouse? What if it was, let's say, in a SharePoint or an extract from a third party system or in a Excel spreadsheet that somebody maintained? Well, then we'd have to figure out a way, okay, how do we bring this into this worksheet? We might be able to bring it back in, but again, those data engineers are the smallest team and the most overworked. So if I need agility to get this solved, this is how that semantic layer starts to creep into the analytics layer. It could be that I just export this raw data into an Excel sheet and my user uses that directly. You can see how fractured and forked this gets. Now let's take a look at what the future workflow could be with this AI disruptor. So we're still looking at building beautiful pipelines, whether it's a data engineer or somebody that's more senior, like a data architect or principal data engineer still needs to understand why we're making these decisions on what a customer is, what a sale is, what's profit. What's the difference between sales revenue and financial revenue? One's for commissions, one's for reporting to the street, whatever. Human beings need to help AI understand what that means. These are the business rules. Once we've got the automation, the analyst is there, and the analyst can decide, well, okay. I want this to be a dashboard and dashboards are very quick and easy to make nowadays. Or I'm gonna build myself a little chatbot, a little talk to data interface and do all the preparations I need to do on that. And now I've got multiple modes of consumption. So my user can look at either one and say, great, I've got my metric or I have a follow-up question. I have a new question. I can go to the AI bot. And if there are new questions, I can go to the AI bot. Now, of course, if there's just completely new data points they need that needs to cycle back through. But now you can see, we don't have so much energy on dashboarding. The engineering effort, particularly when it's accelerated with AI can be faster and we can get more products through and we can get more coverage. If we do have more questions, then we can bring it back to the AI bot. If we have to, we can go back here. But once we build the data product, the use case for how AI can then leverage it replaces the fact that we have to reduplicate, expand, edit, modify, add new filters, parameters, whatever on all of these dashboards. And the good news is, is that the goal, you can see here, is to take the business logic out of this layer and to push it back here so that it's more usable and consolidated. Super important. We'll talk more about that in the semantic webinar. So some things about three point zero. Let's leverage the data platform's power because we can integrate directly with it. We can use its compute versus having it come to us and having to use a separate compute or use our data sets that we've specifically prepared for the analytics layer. We have real time data, streaming data that we can use through these platforms as a result of that. There's plenty of stuff that BI two point zero does really well. It tells stories really well, particularly complex stories. BI three point zero preserves that. AI is everywhere and can do a lot more things though. And it's far more useful AI than the old school ask data, explain data, etcetera. Because the AI is the knowledge worker, now everyone can accelerate how they think about data, how they can think about data preparation, the types of questions and usefulness these tools have for them. It doesn't become haves and have nots. The other thing that's really cool is you can embed these into standard pathways. We use Slack. We can ask questions of AI directly from Slack. Sure you can do the same with Teams and Salesforce has got an SAP and all these other tools are building these so that they you can just do it in your normal place of working versus having to go to a reporting portal or having to go to your data warehouse. The other thing that's cool about three point zero features is these are collaborative, meaning you can share, you can comment, you can you can like, you can subscribe to them. It has very much a social media feel to them and far more collaborative in the way you work as a team. So let me give you some concrete examples. This is the old story dashboard, so selective storytelling. We have filters, so everything that's a bit dynamic, we're gonna put in purple. So as I changed it, my parameters, my filters, whatever, crosstab there might update to reflect it. And maybe my little sidebar there would change too. But I have to click on predefined controls that have to be built to then give me my different ideas. And what happens inevitably is you end up with forty five dashboards, or forty five different filters to try to accommodate all the different things that people might ask, which always ends up to a really subpar user experience. If we take this one step further, we get this idea of number chart paragraph. And the paragraph in the old days, you would take a screenshot of your report, Power BI, put it into PowerPoint, and then have somebody, a BA, lowest rank on the totem, intern, whomever, write a paragraph. Why tell tell executives why this is important. Tell them why that data point's lower than everything else to to give some understanding of what they're looking at. AI can now contextually write this for you because it can see the underlying records, and it's smart enough to start to make some patterns. And, you can help it, you can train it, you can give it context, but the ability for it to write this paragraph, you know, the NCP here is far stronger, more powerful, more potent. Take that a step further. You have the same chart, same sort of starting point, and then we have the ability for you to go and ask questions directly. Why was it like this? And what about this sub domain of this particular product that you're reporting on in those donut charts? Question two, question three, question four. Just go all the way down and just keep building context as you dig in deeper and deeper and deeper. To do that in dashboard land would have taken hours of development cycle or a dozen dashboards to do it or some significantly over engineered dashboard with dashboard actions or synchronized dashboards that you would click through and pass filters through just really complex and overwrought. The other thing that I think BI three point zero has finally come around to is spreadsheets are not going away. I was very much in the Tableau wagon. Obviously, I'm very familiar with Power BI. Most of these organizations like, you know, the spreadsheet's dead, give it up. You should do visualizations. And yet here we are in twenty twenty six with all of finance and a whole bunch of other teams still using the spreadsheet. The thing that's cool about BI three point zero is they have embraced this. They've also embraced the ability to do dynamic sort of ad hoc agile tinkering and playing within a BI three point zero powered spreadsheet that doesn't break the warehouse, but also gives you the agility to start tinkering and do what if analysis. Sigma is a great example of this. What Sigma does is they actually write this back to your warehouse, let's say Snowflake, but they have it in a little separate schema for Sigma. So it doesn't get written back to the data warehouse, which you'd want to have governance and control on. But if you liked it and wanted it, let's say you asked your sales team, Hey, enter in their targets for next year, enter it into the Sigma spreadsheet. They do it. It goes back to the Sigma schema and you want to codify it and bring it into Snowflake, you could. You engineer it over. BI lost the battle of spreadsheets and it's great to see that this functionality is being brought back in because it's useful. It's the same reason people use Google Sheets. Analytics as an interface. Again, building applications, data apps directly into your dashboards. Very powerful. You can leverage generative AI. We use it for some proposal generation. So we'll have opportunities in our dashboard. It'll say, hey, this is the thing that this client wants. We'll say, great. Go look at our context layer that's out on Box and generate me a proposal that's really close to what I'm looking for. And it might get it perfect. It might get it within eighty percent, but it gives me a huge starting advantage. Our colors, our language, our service slides, drawing from Salesforce to populate costs and estimates and all those sorts of thing. All I have to do is just go make sure it looks sense and and make sure it tells the story. A massive accelerator. And these are the types of things other people are doing too. Speed to insights, so again, ETL tools, analytics dashboard preparation, data preparation, all of these things are becoming faster and faster and faster. Before I let you go, I've got a little thing to say how we can help, but I just wanna give you guys one really, really, really important thing. I would have said maybe even three months ago that your data is the most important commodity you own, but I I need to be more specific than that. Your data lake is not the most important thing you own. It's your business logic. It needs your data lake to be useful, but your business logic is the essential thing that defines how you guys think about your business, how you think about your customers, the decisions that you've made over the collective experience of your organization being an organization, that's represented in your business logic. Do not give that away. There are a lot of organizations that are making a play to own your business logic. Oh, it's convenient. You get these embedded AI agents. This language is really easy. It's much easier than SQL. The problem is is that if you lock your language in, you are then tied to that customer, tied to that vendor rather, that platform until you decide to reprogram it yourself to get out of it. And I will tell you, having done this for a lot of customers, that is quite a painful and expensive process. This is one of the reasons why we love tools like Snowflake because it's SQL, it's iceberg tables, it's extensible. It integrates with another tool like DBT beautifully. Own your business logic. Do not tether yourself to a particular vendor. There are consequences when you do, and that inevitably is cost and agility and flexibility. Off my soapbox. I'll I I will that will not be the last time I say that, but I keep seeing organizations trapped and I'm like, it's gonna cost you seven figures to get out of this application because you've got twenty years of stuff that you've locked in here. So we can help. We can help in a lot of ways. Hopefully, you remember the beginning of this presentation, whether it's strategy or platforms or AI or data or pipelines, or us supporting those or supporting your community, building analytics for you, all sorts of advanced stuff we can help you with. One thing I'll call out specifically that you might think about is when you are ready to choose any analytics tool, your BI three point o tool, or perhaps you're looking at building a better data platform. And a lot of these data platforms like Snowflake have really cool analytics features in them, like co work, where you can do natural language directly in it. Those things are really useful, but I would be cautious about trying to navigate it on your own. There are a lot of things that you need to think about from features, roadmap, pricing, your use cases, those types of things that I think it's professional like us could really help you with and save you time and energy in doing it. So we can help. We can help you do the spec and select process, whether it's for the data, analytics, governance, whatever. And there's quite frankly, a whole bunch of other stuff we can help you with. The best way to find out how we can help you is to just contact us. Scan this code, go to interworks dot com, respond to any of the webinar information that we have sent. And one of our salespeople or me, myself included, happy to jump on and just give you some real honest dealing in terms of what's possible, what the path forward might look like. We are not here to win a deal. We are here to win a relationship. We are looking to work with you for the long term, and we don't do that by trying to cash out. So we're not here to sell you a big tool because eventually, you're gonna say, why'd you sell me this tool? We're here to build value with you for the long term. Most of our customers we've had for years and years and years. We're committed to your success. Scan this. Contact us today. Happy to help. Okay. We've got about five minutes or so. There was a lot in there. So if you've got any questions Oh, I just saw the little boy did the coding. I did some coding back in the day. It was all basic, like literally the programming language basic. I did some Pascal and some C in there, I think, before I put those lives behind me, cold fusion. If you've got any questions, drop them into the chat. I'm happy to answer any questions that you might have. Otherwise, if it's quiet, I might let you guys off the hook five minutes early. See what you got. For those folks that registered, hopefully, that's everybody. We'll send out an email probably in the next day or two saying the recording's ready. So if you wanted to see this or share this with a colleague, you'll have the recording, and it's just on the interworks dot com website. Doesn't look like any well, we've got a hand raised. So, Hanin, if you wanna just drop your questions straight into the, chat, it is easier than me trying to figure out how to unmute you, which I don't know if I even am allowed to, like with the permissions on these webinars. Oh, sorry. By mistake. No worries. Well, I'll let you guys go there. Thank you so much for joining. Don't forget we've got a webinar every single week covering topics like this, forward thinking AI, and how we get the most out of our data. Pleasure chatting with all of you. Let us know if we can help. Thanks, everybody. Bye.