Alrighty. So let's just get kicked off. I am going to run this show solo, but I have friends that are watching, letting me know if they could see my screen. I will warn you, it is cheating in Melbourne, which makes me scared that we're gonna lose power or something. If we do, stand tight. We'll see if we can't restore services, But it's it's cats and dogs here in Melbourne. Alright. So we're gonna talk about the future of analytics, specifically that journey to BI three point o. If you don't know what BI three point o is, don't worry. We're gonna bring everybody up to speed on what exactly that means and sort of the iterations of business intelligence until now. I'm Robert Curtis. I'm the managing director for Interworks. I'm based out of Mobrin. I get to look after Asia Pacific. I've been with Interworks for going on in total about twenty years over a couple different stints. I started when I was just a wee lad back in university somewhere, I think two thousand. So I've been here for a very long time. It's been my pleasure and honor to build our operations over here in Asia Pacific. I'm just seeing it doesn't sound like I have an echo. Maybe someone does. Audio coming in fine. Alright. If it does sound funky, let me know and I'll see if I can fix it. But I think that might just be one particular person. Let's do a little bit of an introduction into Interworks. So we are a data company and we help with data strategy solutions and support. So really what we're thinking about is sort of that lifetime end to end journey of data within your organization. What do you want to do with it? Just remember, I probably need to put some background noise on. Otherwise my headset will shut off. This has happened to me. For those folks that have been on here more than once, you can remember the times it does happen. I'm just going to put some dulcet tones of Neil Diamond in my headset. You won't hear it, but I will. Data strategy solutions and support. So again, if you are thinking big ideas of what you want data to do for you, and that might be in terms of how you protect it and govern it, it might mean how you use AI to drive value in automation. It might be building the solutions to do these things. It might be analytics or your data warehouse or data pipelines, or it might be supporting your organization, your community, your individual users, or your applications themselves. We do all of that. We don't do too much outside of that, but because we have such a narrow focus, specifically on sort of that niche of data, we've been able to build a tremendous amount of pedigree. More specifically, this is sort of a life cycle of the things we do. So again, starting with strategy, working our way through helping you with foundations with data and governance and platforms. Platforms could also be cloud, as you can see in the icon. Certainly probably is cloud nowadays. Solving problems and adding value to analytics, AI, data science, as well as a host of other things that your imagination might sort of conjure up as a combination of all these things. And then again, as you get going, building that community, adding skills to individual users or groups of cohorts, and then supporting applications as needed. We can do some of this or we can do all of it. We are literally happy to be a Swiss army knife or a puzzle piece to make you the organization and the data team that is the most effective. I did mention we've been doing this for a long time, so we do have some, celebrations. I pulled this slide out of an older deck, and I'm realizing now it's thirty years as of this year. We celebrate our thirtieth anniversary in October of two thousand and twenty six. Coincidentally, it's also the year I celebrate a milestone, which is fifty. I am old and should be retired very soon. Put out to pasture, past my prime. We have a lot of customers that we get to work with. Seventy five I think it's seventy seven now of the Fortune one hundred are or have been IntoWorks customers. That's an amazing, stat for a company that's probably three hundred folks worldwide. Again, we are specialists. We focus on excellence, white glove service. We are not we have never intended on being the biggest. We simply just wanna be the best. We have a blog, and certainly, you'll see some of the ideas that I explore here detailed in a lot more granularity and more content, on our website. And we get somewhere around three point five to four million page views every year for our interworks dot com blog. A lot of the people that I get to meet know me because of our blog. Oh, you guys are into works. I read how to do dashboards for you. I learned how to do, ETL or functions or or whatever the numerous things that we do. We have thousands of customers worldwide across every vertical and every site you can imagine. We work with small organizations all the way up to the multinational with hundreds of thousands of employees. We don't pick a particular type of customer. We like to pick a particular type of person, somebody that is interested in thought leadership and true partnership and a long term relationship. That's where we do our best work, and those are the types of people we like working with. We've got a lot of awards throughout the years. Think somewhere around twenty to twenty five partner of the year awards across different partnerships and data analytics, ETL, governance, etcetera. But one of the ones we're most proud about is a few years ago, Forbes compiled a list of small giants, small organizations across different verticals that do amazing things, we were picked. Interworks was picked as one of them. So again, less than three hundred, but making a global impact with some of the largest companies you can imagine. Let's get started. That's enough of the preamble. You might be wondering why I have a little face here of a child. Well, we're telling stories. Right? Just like we would with a dashboard. I'm here to entertain as well as illuminate. And this is our BI journey until now. You know, the old man there, middle aged man, the youngish, but perhaps not super young man, and then the toddler. Now you might be saying, wait a second. You said this is BI three point o, but there are four dudes standing there. What is going on? Well, I'll explain. But let's start with some trivia. This is gonna be personally embarrassing, but I was working on this last night. And when you're building a deck at eleven PM at night, some ideas sound better. Less less better than in the light of day. So here's my trivia question for you. I'm gonna show you two images. And if you feel brave, chuck those into the chat as a quick answer. But what do these two things have in common? One is this old looking computer with this peculiar software on it. And the other one is a picture of this lad. What do those two things have in common? Any guesses? I kind of hinted at it at the start of this presentation. See if anyone has any ideas. Age, data, fifty. Old. Okay. Alright. Alright. Maybe maybe this was a mistake. I'm starting to feel a little bit persecuted. If you said that both of these items, and this is me over here with the cowboy hat, that's me all the way back in Oklahoma when I was about one or two years old, both of these things are from the seventies. Very old. But it's a way to illustrate, for those younger folks on the call, things before two thousand are still valuable. There were good things that happened before then. I am from the seventies. I was born in seventy six. I guess this picture of this old man is actually representing me as well as what we're gonna talk about. But more importantly, that computer with that application on it was the very first BI tool known VisiCalc. It was released in nineteen seventy nine, and it was essential in in justifying the purchase of PCs for business or for the home user. That killer app is what people were looking for. We've got the technology, but we don't know. We don't have a use case. It kind of sounds like AI, maybe even just twelve months ago. We've got the technology. We've got LLMs and vector databases, but outside of using it to create custom images, generative AI to build images from my kid's school report, what exactly are we doing with it? VisiCalc was the killer app that put the PC on every desk throughout the world. And what VisiCalc did was it was the first electronic spreadsheet. And this is sort of your pre BI, your pre one point zero wave. So let's call it pre BI or BI zero point zero, but this is the old school. And for some of those folks, the some of the folks on this call with us, those logos will will not be from a an archive. It'll be from your living memory. That's VisiCalc. That's the first Excel, which was in eighty five. Lotus one, three, I think was before Excel by maybe a couple years. But this is where computers started to have hard drives and I think in the sixties, they were able to stay store storage. They were able to store data. That helps if I don't flip those words around. And start doing processing there on the hard drive. Again, we're talking sixteen mega RAM or sixteen k of RAM, and then thirty two for Apple two point o, which is where a lot of these tools really kinda got popular in the home. But this VisiCalc and some of these other tools that followed it were transformational because a lot of these spread sheets had to manually be done. Manual calculations, manual modeling and forecasting. And this this spreadsheet, green and black on those old CRT monitors, automated a lot of that. And it gave you the ability to do it right from the screen. Now, of course, when you wanted to do something with it, you'd go print it out on a dot matrix and then share it around the office. Hey, look at these numbers. But it suddenly made data accessible for the first time. And in a way that was representative of something people were familiar with, which were the accountant spreadsheets. And of course, that's how data came first to the accounts, to the finance people. And slowly, you'll see it's been opening and opening and opening up. Now, of course, there are massive limitations. Could book in this as part of BI Wave one point zero, but it's so nascent, it's so foundational. I think it's good to call it prehistory. And the limitations are exactly what you could imagine. Almost no visualization capabilities, very, very limited formulas and customization, almost no data preparation, no ability to check for data quality. Performance was horrific. So again, you can imagine trying to do big data on Excel today. Well, imagine trying to do something like that on these machines back then. So this is really just that first step to unlocking people's ability explore and interrogate data. What's even more exciting is the concepts of this came back from like, there was a guy that wrote a book on sort of using banking and other businesses back in the 1860s. Again, no computers, just this is the analysis and this is what people are doing. And then I think the very concept of business intelligence was like in nineteen sixty five, so it took years before technology could catch up. But in nineteen seventy nine, it did. Now, I've talked about BI operating models quite a bit. And you can go, I think, about a year and a half back and we did a whole webinar on it. But I think it's interesting to kind of trace these early BI waves to what was the most prevalent BI operating model at the time. And for BI zero point zero, it was very much the decentralized. You've got a couple of people with this stuff on their computer, and they are building cool stuff, printing it out, and showing it to everyone around. That sort of early evangelism stage was as best as people were going to be able to do here because there was no governance, there was no centralization of data, there was no centralization of anything. It was individuals doing their best to automate processes that was normally written out by hand. So, again, from a diagrammatic point of view, you've got some people sort of spread randomly without the organization doing their best to try to leverage harness data to make decisions. So that gets us up to BI one point zero. And now we get to see objects that are literally, it's kind of a pun, I guess. You get to see applications that are still lingering around, Excel's around, but it's very different from the days of when it first came out in eighty five. But these are things that are still there. BusinessObjects, Cosmos, MicroStrategy. And the thing that made this really possible was the theoretical concept of a data warehouse, which is let's get everything in one place. And the first commercially available version of that, I think was in 'ninety two, which was Teradata. Now there were other things that were happening prior to that. So you could say, well, where do you want to really draw the line? We'll draw the line at Teradata, released for public consumption. That also gave us the ability of facts and DIMMs and OLAP and building predefined reports. You can schedule these things so that doesn't require all the automation. There's an abstraction layer, AKA a semantic layer. You can put your business definitions in there. And as a result of all these big, expensive, centralized things, it made sense that IT said, you know what, let us take care of this. Because there's a lot of money going into this. It's very complex. Nobody knows how to use it. We'll learn how to use it and we'll make it work. But it did allow us greater complexity with data, greater analysis, bigger questions that we're trying to solve. Now, of course, these are large machines, very expensive. So the limitations are pretty obvious, a lack of agility and speed. As a result, there were really long queues for people to get the reports they needed if they got them at all. Again, the solution was to try to solve it in some sort of a spreadsheet. That sounds familiar. You could only do batch data. You could only do structured data. It was very costly and there were very high barriers organizationally to do this, as well as individually to learn how to do it. But we were on our way. From a operating model BI one point zero basically created the report factory. Everything was expensive. Everything was very complex and technical. So everything got centralized. And a lot of organizations have stayed there. And for some, it is a deterrent that they need to evolve. And for others, it makes total sense. For instance, if you're in financial services, having a highly centralized approach makes good sense because there's regulatory risk if you are not a bit more regulated and controlled internally. For other folks, or maybe even other departments, it might be good to sort of give them a bit more latitude to innovate. Again, that's flavors for different folks. And what does that look like when we actually lay over an operating model? Well, it means you've got all your your data workers in one spot, going through the data sources that they need to work, getting everything nice, and then building dashboards for people. So rather than individuals split throughout the organization trying to hack it together on their desk, Now you've got a team of people, the Avengers, building these reports for other people. And of course, you can see as a result, you've got business users that are being underserved because they're waiting for IT people to translate their needs of their business back to them. And sometimes that works. Meaning I, the IT person, understand what you need and I get your report eventually that is helpful. Or it could be you are so specific on what you need, and I don't understand that particular aspect of our business or those metrics, I'm gonna give you something that I'm going to have to fail at, fail at, fail at until I get you something that's a little bit more useful. Or maybe our data doesn't even support what kind of questions you're asking. Again, that translation from user to IT person is a bridge. And oftentimes that bridge is pretty difficult to cross. This gets us to where we were just recently. In fact, a lot of you probably are still here, which is BI two point zero. So we're now at our grown person who is not too old and not very old like Rob. And these are tools like Tableau, Power BI, Qlik, etcetera, etcetera. Tableau is the revolutionary component here. And not saying if you see a tool here that it hasn't continued to evolve. I'm just saying these are the tools that became the game changing thing here. Tableau was the very first thing that said data for the people, let's democratize data, let's get regular business users solving complex problems, which unloaded a whole bunch of opportunity and revolutionized the way that businesses think about data, use data, and the culture that they're trying to build. It is pretty universal for businesses to say our most important asset is probably a 1A and 1B. 1A is our people, but 1B is our data. And those two things are intrinsically related. So with the tools of yesterday, not yesteryear, but yesterday, and maybe today, you have the ability to analyze a lot of different type of data. So batch near real time, near near real time, you can start to do with more advanced data structures. You can do semi structured, unstructured. You can do advanced use cases. You can start to explore machine learning data science. The visual part of this is much more powerful. You can build interactivity into these dashboards. You can do self-service dashboards. Your business users can use dashboards. It is mobile friendly, so you can put dashboards on your phone. You see a theme here? And then there's just starting to get some baby steps of AI and how they help you with dashboards. Now, if you heard the word dashboard, that was intentional. Because this is probably the big thing here that has really spun people into say, you know, we need something better. And there's a lot of organizations, a lot of tools that have specifically addressed this dashboard issue from a marketing perspective. And that's BI two point zero said everybody can build dashboards. And so dashboards by default became sort of the common currency of insights. The lingua franca of data, meaning every question can be answered with a dashboard. And there are some folks on here that if we were to have an open format saying, who here has over a thousand dashboards in their organization? A lot of US would raise your hand. If I were to say who has ten thousand organization, there'd be people raising their hands. We have clients that are dealing with that right now in a way to sort of rationalize those down. That is the problem with BI two point zero. Everybody can create a dashboard and so everybody creates dashboards. Then the question becomes, if there's an explosion of dashboards, how do I actually know where to go get my insight from? If I have ten thousand choices, does that mean I have no choices? That's part of the trouble. With that dashboard explosion then becomes the management of them. So now, overwhelmingly, you're probably seeing analytics teams that are really heavy on people building dashboards and supporting dashboards. And the people behind the scenes building the data, they get the skeleton crew. That's a bottleneck. Report management, building dashboards, people trying to find them, organizing them. And again, because now we have so many analysts and so few data engineers, your business logic starts to leave the data warehouse, that abstraction layer, and it starts to enter into your dashboards. I have seen organizations that have hundreds of line of custom SQL in their Tableau workbooks to try to make up for the fact they don't have a curated data source in their warehouse or their data mart or their lake house, whatever terminology you like. So the challenge then is while we have a lot more power and flexibility, we're inheriting different problems that end up with the same results, which is users are still struggling to get the answers in a way that they need in a way that's helpful. Now with BI two point zero, you can have all of the previous versions of this that we've seen so far, but this is generally where you start to see people arrive at maturity, which is this federated idea. The central's gonna do some stuff. Largely, they're gonna prepare data sources for us, and they might build the most operational, most useful reports. And then we, the business units, are gonna have our own ability to do stuff. And maybe if we're data mesh or we've got really, really interesting data, like for marketing or whatever, we'll build those data sources and share those out for everyone else too. But largely when we're talking about Federated, we're really talking about the analytics part. Everybody can do their own analytics and we're gonna support them at the departmental level or team level. They have the ability to drive innovation quite quickly. And, ideally, they are capturing the business logic in a way that is really useful for them. Now there's challenges, of course. We're not gonna go into every potential ramification of how this can get complex. But this is what the diagram looks like. Now you'll notice we've got data workers inside the business units because these BI two point zero tools are much easier to achieve mastery on. Power BI, Tableau, etcetera. IT is still thinking IT or centralized data services, whatever it is for your organization, you can kind of fill in the blank there. They still worry about the big data sources and and getting them together and building the warehouse and hopefully building that curated data layer. But there's still queues. There are queues within each department because, again, not everybody in BI two point o can be be a data expert. It's impossible. Some people, we don't want them to spend the time and energy becoming expert. Like for instance, does our CEO need to be able to do self-service with data, and dashboards? Probably not. It's probably not a good use of that person's time. So there are still queues. They're just distributed. And there's obviously still a huge bottleneck on data. So federated BI, is it self-service? Well, it's far closer to self-service than anything so far. So we line these up side by side by side. We've got PreBI, which is the fifties to the eighties. Again, we're building foundational things. It's explosion of the PC, the exponential curve of computing power and hardware and all of the things that sort of empower the foundational bits that are yet to come like cloud, basic data capture processing. Again, with serious limitations, we are really talking about Proto BI here. But the thing was, hey, we can do stuff. I can look at this spreadsheet. I can actually start to do some automation here. We get to wave one, that's nineties and two thousands. Here we can aggregate our data and we can present findings and we can ask what happened? When did it happen? We made these sales last year at this quarter. Again, very centralized, highly governed, which is a big plus. IT generally owns all of it. It's full service. And the concept that really empowered this is this idea of centralizing your data in one place where we can benefit from it. There are still organizations that struggle on that centralization part. Again, still takes time, still takes money. Wave two. This is where we have been for about the last, say, fifteen years. But it gives us more ability. We can explore, we can predict, we can analyze, we can ask more questions like how did this happen? Why did this happen? It is business unit driven versus IT driven. So long as they have good data, of course. It is the ability to visualize and tell stories versus rows and columns of numbers, which then takes the human effort to understand what's important about it. I'll say self-service, but we might think about self-service maybe at the departmental or team level. There's always somebody. Susie at that desk, she's really good at Power BI. So whenever I have questions, I just ask her and she helps me out. So, quote unquote, is that self-service? Close. But but not exactly. But the thing that's really good about data analytics with Wave two is not only are we getting foundational stuff, we're getting data that we can use and trust, but now we're able to ask, we're able to get better answers. Not just when it happened, but why did it happen? You can look for outliers. You can look for the depth, the stuff that's actually going to make some decisions, actionable insights, which brings us to wave three. And we're here. We're in wave three right now. Now it is the beginning of the wave. There's going be a lot more cool stuff to come, but we can already see tools in the marketplace at maturity. And they'll only continue to get better. So let's ask why we have a wave two coming, wave three coming rather, and then we'll tell you exactly what we're seeing. So these are the big trends right now. The biggest one is the preeminence of the cloud data platform. If you make one decision in your life, it is probably what, where you are going to put your data. Because right now the race is for a few data platforms to eat up everything that you need around it. So that's Snowflake, that's Databricks, that's even say BigQuery. They're all saying, not only can we be your warehouse, we'll be your lake, we'll be your AI, we'll be your governance, we'll be your transformation. We can do all of these things for you in a single platform, and we'll do it better than the previous generation of on prem data warehouses and databases. And they're right. They are doing it better. There's different flavors. So if you if you have, a particular type of organization, you might lean into Snowflake. If you've got another type, then you might lean into something else. But this is probably the most important decision you make. We don't make decisions off of what you're gonna do based off of the transformation tools or how you're going to ingest them. Those things are use case by use case. You make decisions based off of your big platform. If if you think about it as a marriage, this is where you're putting your ring. That's super important. And every tool, every analytics tool needs to integrate and leverage the power that is growing in those platforms. And they're only gonna get even better. They've exploded in terms of the utility in the last twenty four months. AI is being integrated into everything. It is making your analysis more powerful. It's making the storytelling more powerful. It's making your user experience more powerful. One of the big things I'm talking about is conversational analytics, my ability to talk to data through natural language. AI can also provide proactive information as well as contextualizing the data that I'm looking at. AI can augment the workflow so that data engineers can work faster, and there's several ways that that can happen. And also, my analysts can prepare data faster so that we can, again, unbottleneck that critical step in it where we need to get curated data, curated data for users, and now curated data for AI. All of this stuff integrates within the stack with other tools. And all of these things, particularly your new analytics, your BI three point zero tools can drive actions from the interface. We'll talk more about that here in a second. And again, the whole idea is to try to intersect data and the value of it within the way human beings think and the way the human beings in your organization work. We want data to find them. We want insights to find them versus them scratching their head and hoping somebody can come along and help them. Let's talk a little bit more about it. So BI three point zero. This is AI serving up insights, and it can it can do that in a lot of different ways. So again, we talked about cloud data warehouses, AI, but one of the things we really wanna talk about is it's not just the dashboard anymore. We really need to embrace all of the ways that people work with data, which means new ways. But also looking back, there clear pathways that people have established. This is how I like to work with data. This is a valuable way. You cannot throw that away like we tried to do in BI two point zero. I'll give you a specific example in a second. The other thing is there are much lower barriers to insights when AI can be the technical resource that's dedicated to every user. And then analytics, again, as an application interface. Let's let that be an action layer, not just a presentation layer. The limitation here is until AI figures out a way to help us build semantic layers, it can certainly accelerate us now, but until it builds them on their own, human beings need to spend a lot of energy on building curated data sets, which then unlock all the other things. Your guess is as good as mine as to when AI can come and look at some disparate data sources, join them up, validate them, de dupe them, integrate them, transform them, and then polish them. We're we're still a bit of a ways off, I think. So when we bring everything back together, this is what we look like. So that wave three, it's available now. Again, anticipating things, enriching decision making, enriching the context. You can answer all of the questions that you might think about asking of your data. AI augments your your ability to build things. It augments your user's ability to interact with There's smart automation. You can personalize this very, very acutely in a way that wouldn't be possible without building ten thousand dashboards. And not only are you getting better answers now, because you can interact with the agent directly, you're getting better questions as well. So just imagine having a fully trained, highly skilled IT person with all of the skills and understanding of your data and your analytical sitting at your desk waiting for you to ask questions. That's what AI, agentic AI, all of this stuff. That's what it's promising. And there's opportunity to leverage that today. And again, early investment means you're going to get benefits now, but bigger and more exciting features in these tools are on their way. So let's go look at what federated BI look like. So we looked at this slide already, and this is us having departmental power users sort of helping support less technical users. And this is about as far as we could get today by the limitation of our tool. Queues are shorter by department, but they're still there. And data is of course, the critical piece that makes all of this work. By us switching over to BI three point zero, this is how the ecosystem changes. Rather than us having dedicated power users or data workers in the business unit, let's bring everything into a centralized team to build as much amazing data as possible. Because we don't need to have them buddy buddied up with nontechnical users because that's what AI is going to do. AI is going to be their window into data, their guide into seeking insights and actionable intelligence. And as a result of us no longer having dashboards as the common currency, the unit of measure of how we generate insights, we don't have these cues anymore. I'm not saying you won't have dashboards. You will not have ten thousand dashboards. Somebody wrote in here, we have four thousand objects in Cognos and another two thousand reports in Power BI and a whole bunch of stuff. Just lots and lots of stuff. And if you actually got the usage statistics of those, a fractional amount would actually be used on a monthly basis. So we don't need them. Let's get rid of them. Let's keep the fifty or so really, really awesome, very complex reports that would be hard for AI to generate on the fly and give everybody else the ability to ask data. So let's do a workflow. This is what it looked like before. So we have our data, and that might be multiple systems, but we've got that data coming in. We're going to integrate it, clean it, validate it, curate it, and then we're going to build it into these nice little packaged curated data sets. So the manipulation of that data is quite manual. But once we get it in there, then we can build automation and orchestration to get that curated data set ready to go. And a data engineer physically does that. Then we have to do something with it. So let's put it into a dashboard. And again, traditionally that has been a very manual exercise done by an analyst. The analysts are closer to the business. A lot of them sit in the business unit, so they'll understand the business logic that much better. We then give that to the consumer and say, hey, we hope you find valuable things out of here to make your management of your area of our business better. So hopefully, they get little stars and and find useful stuff. If they have additional questions, like, you know, there's something interesting in this dashboard, and I have another question about it. Or this dashboard takes me half the way. I need I need more data to actually have good insights. The problem here is that it then triggers a very manual sort of a manual refactoring, or we have to go back and look at the whole process. And we might go all the way back to the dataset here and say, well, we don't have that data. So we've got to go back to the start. Or we might jump over here and say, okay, we've got to go and re prepare that data. Or we might have to redo the dashboard. But whatever it is, it becomes all the way back to a human being having to sort of try to fix things. And if we are trying to analyze data that the data engineers have not gotten back to yet, or haven't even started on, then we have this raw clunky data over here. And I see this in organizations where analysts are like, I have got to solve a problem, and I need that data that we haven't gotten ready yet. Or it's coming from a source system or it's coming from a SaaS application that we have access to, but it's not in our warehouse, it's not curated, I can't wait. Which means your analyst is probably doing their best to try to build what they can in their dashboard, which means their business logic is going in the dashboard. Or even worse, you've got somebody that's not technical that has to solve that problem. And they might be pulling data out of financial systems. They might be pulling out of marketing systems and they are doing spreadsheet analysis to try to solve that problem and get that insight. And that's even worse because that is very manual and all of that data is ungoverned, all of that business logic is siloed. If they are doing really sensitive information in a spreadsheet, you have no idea. So that is a huge risk. And also, we are potentially not potentially. We are one hundred percent duplicating effort across the organization because I guarantee you people are building similar things that they don't need to be building. This is the old way, or rather this is as of today. And what I'm proposing to you with BI three point zero is that we could think about our workflows in a different way. We still need data engineers as of right now to help us put business logic in and make sure we do all the things that are appropriate with it. Right? Enrich it, dedupe it. AI is not there yet. And great, if it is happy to completely redo this workflow. But we do need to get as many curated ready to go data sources as possible. Now, on the other side of that, we'll have analysts, but hopefully we don't need as many. We'd prefer to push the analysts over into engineering. But what the analysts now can do with modern tools, BI three point o tools, is they can choose. We can use AI to help us automate building dashboards so that they're built faster and maintained better. Or we don't think a dashboard is the perfect thing here. Let's build a little AI agent, which can also be automated because it's sitting on this beautifully curated dataset that our users can interact with. And our user gets the benefit of both. And they're like, great, Amazing. But what if they have more questions? Well, they can go and ask the AI agent. Because again, we're not building something in stone. We're building somebody that you can ask. And the AI agent fills that role. It is somebody right on your laptop that you are building so that your marketing people, your salespeople, your managers, whatever can say, hey, what about this? What about this? And what's cool about these BI three point o tools is they will capture the queries that people are asking. And that becomes a feedback loop. If the analyst says, you know, we're getting a lot of questions about this thing. It's time to add our twenty sixth dashboard rather than our ten thousand and one. Or we can go and modify dashboard. You know, we can clarify this because we're getting enough questions about this. Or we can go and look at more data because we're getting these questions. But that cleans this up so that there is the ability for users to get more direct feedback and more direct action, more support in an automated way. So when we think about what three point zero can be as a list of features, I'll give you a summary and then we'll go into some specific ones. Again, the data platform, super powerful. Let's do everything we can to take advantage of all of those things that are being built into a, you know, hypothetically, infinitely scalable cloud with all the different workloads that they can build on there. Let's do everything we can to take advantage of that. We can go as near real time data as you can because, again, these new BI tools are sitting so close to the data that we don't need to put inhibitors in place. If there's things that work in BI two point zero, let's keep them. We don't need to get rid of dashboards just for marketing. Dashboards work, but they don't work for everything. And that's an important distinction. Let's put AI everywhere we can where it is safe, effective, and efficient. And as AI becomes more safe, more effective, and more efficient, we're going put it more places. Let's accelerate data readiness. So let's redistribute our teams so that the folks that are spending a lot of time building their semantic layer in data preparation instead are building it at the centralized warehouse. Let's make the tool essential for everyone. So again, let's put it in the pathway in which I work. Don't make me go look. Said, what's the link for the Power BI reports? And I thought we have some Tableau stuff and, oh my gosh, we've got some embedded reports with this application. I can't remember which one I need to go to or which one's right. And again, the other thing about three point zero is it can be very collaborative. There's shared ways of working for those folks that have been in BI for a decade, you know how painful that has been. And then the ability to add a social function to it, likes, shares, all those things that sort of upvote or or or continue a conversation inside of these reports. So let's look at some specific examples. So the dashboard. There are really good things that happen in here. We have filters, we have parameters, we have dashboard actions, all kinds of cool stuff in there. Let's keep it, but only use it where needed. The dashboard, I would say, is for selective storytelling. If it's there to solve a problem or there to satiate a self-service need for a broad nontechnical user base, we could probably make this experience better. But if we've got complex ways of thinking about particular metrics that we have ironed out over a long time with a lot of really smart people, I'm not going to be able to ask AI that question with that level of specificity. I'm gonna need a lot of people thinking a lot of deep thoughts about the business and embedding that logic into a tool like this. Great. Let's use it. No reason to throw it out if it's useful. The next iteration. This this idea of a number chart paragraph. Think about this as like a a stepping stone. A lot of times I get from users, okay, I see a dashboard. I don't understand why it's important. What's important on this chart? AI can help contextualize what you're looking at. So it can write and say, hey. This outlier here is really interesting, and we sold a lot of couches this month. Isn't that weird? And we sold a lot of couches this month because this salesperson did an amazing job. You know, whatever the context might be. We put some essential sort of things there, and then we add context to it. And if you change the filters, the context changes with it. A lot of times what I see organizations having done over the last ten years, screenshot me that Power BI report, and then you go have one of your people write what it means, and then we'll put it into a PowerPoint. If we don't need to do that, amazing. We then get to another level deeper, which is let's embed the actual agent inside the dashboard. Maybe there's some context, and then we can go ask questions. Why did it do this? What about this? If we changed our prices by this much, what what what might happen? As AI gets smarter and your data gets richer, those questions are gonna be more valuable. And and there are also personas that you can start to apply to these agents. Answer me as if I'm the CEO. Answer me as if I'm a product manager, and I'll attempt to try to give you those sorts of perspectives. High level trends versus the details. AI is already doing all of this stuff right now inside these tools. Some other things that are worth one. As a long time Tableau user, as a trainer for Tableau, as somebody that taught people how to use Tableau all over Southeast Asia, I have to confess we lost the war. Whole, one of the big things that we're trying to do is move people off of spreadsheets onto Tableau. And quite frankly, they won. I had webinars from like maybe five years ago, how to change a crosstab culture. Well, the reality is, is that there's a very good reason people use Excel. And for these types of analysis, they really wanna get their hands dirty. They wanna they wanna calculate on the fly. They don't wanna write a calculation. They wanna put it in a column. They wanna try different numbers and see if I do this gap or I put this target or I put this historical, whatever. I wanna see what it looks like now. And I wanna play with things and do some analysis. Excel is still the number one most used BI tool in the world. It just so happens to be a dataset at the same time. And together, that's reversible. Why are we fighting this battle? Let's bring that workload in. Let's bring the finance bros into our analytics tool in a way that we can give them great performance. Instead of a million records, we can give them four hundred million records. We can give them the ease of use that these new tools have, better wizards, better training, better understanding more quickly. Give them all the advanced features that AI tools have, as well as the ability to have a spreadsheet that they're working with and then do all the visualization right next to it. And also because we're bringing it into a new BI three point zero tool, we can finally add governance to this. We can capture the logic that they're putting in there. We can automate the logic for them. We can make sure the data is being used in a way that is safe and appropriate. Analytics as an interface. This was a big light bulb moment for me about midway through last year. I was thinking, okay, BI three point o, you got AI, you got some new ways of doing things. Great. I got it. And then someone said, no. You could actually take this this analytics tool, quote, unquote, analytics tool, and overlay it directly into the features and cool stuff that your data warehouse can do. So for instance, here's a dashboard, and maybe it's an interworks when we're looking at all of our leads and all of our business stuff. Click here and you can generate a proposal from the dashboard. Click here, you can generate some drafts of an email that you can go proof and then go send. All of that stuff is capable for all of you too. Snowflake, Databricks, BigQuery, all of those things have generative AI features in there. And these new BI three point zero tools sit right on top of them, and they can bring that stuff to the surface. It is extremely powerful. In fact, I would say outside of the AI bit, this is probably the number one thing that has got us most excited. And as crusty old long gray beards, like I showed you at the beginning of this presentation, we don't get excited about much anymore, but this I'm like, wow, that is a new paradigm. That's exciting. And of course the speed to insight, not only is ETL becoming faster and the building of dashboards becoming easier, but the ability to do data preparation is easier, faster, more intuitive. So again, we can do more things from the analytics space. It doesn't mean we want to build the semantic layer there. I want to be clear. We want the semantic layer to sit in the data warehouse. But it's doing that last mile bespoke, enriching, changing, whatever it is that you need to do for that specific iteration of that data set to solve your problem or to tell that story. You can do it easier. I do wanna give you a word of caution because it is early days for the industry. I think there are a lot of really exciting tools out there, and there's a lot of tools from BI two point o that are positioning themselves as three point o. And some of these will be amazing tools and some of these won't. That's just the way things go. So I'm gonna give you guys some words of caution. I would say the first thing that you need to choose is your data platform. Don't start with your analytics tool. I know in the old days, that sounds like heritage or sacrilege, because you could say, well, Tableau is tool agnostic, you should start with Tableau. And in those days, you were right. You can pick an analytics tool and it will make any combination of things behind it work. But now because there's so much power in your data platform, you want to leverage that and your analytics tool sits on top of it, not next to it. It's the question that everyone has been asking since the dawn of time. Should I go full stack or should I go what's best in class? There is a place for full stack. It might be cost ramifications. It might be a lot of different things. I think there's a really good case for best in class as well. That is for you to decide, but I think you should make that decision versus fall into it. This is a big one. Should you be pre paying a premium for BI three point o features in your analytics tool? I don't think you should. I think that is now the expectation that the market provides us BI three point o features for two point o prices. If someone says here's our BI two point o tool, but if you want that sexy, sexy three point o stuff, you're gonna empty the pockets. That is not true anymore. I strongly advocate for you to still look at the same sorts of budgets that you would in the past. BI two point zero was expensive enough. The other thing I would recommend is when you are selecting a BI three point zero tool, have a clear idea of your ways of working. So the the data life cycle, your your data DevOps, your BI operating model, and choose a tool that matches that. Don't choose a tool that you then have to reengineer everything around. Again, remember, the center of gravity is at the data platform. It's not at the analytics tool. The analytics tools and your ETL tools, for instance, need to conform to the way of working that you've set up by the big philosophical decision you've made about what platform, what data platform you're using. And if there's nothing else you take away from this, this entire conversation, in fact, every webinar I've ever done, it's this final thought. Your most important, most valuable asset that you own is your business logic. It is your data combined with all of the special stuff that you put on top of it that defines the way you guys do business, the way you guys make decisions, the way you guys have recorded all your events transactions. All of those things that are built into your data, your curated datasets are gold. Do not give them away. Do not lock them into proprietary language or vendor lock in. Do everything you can to keep them one hundred percent yours and extensible. And I say that to my disadvantage because we make a lot of money helping people dig themselves out of proprietary software so that they can actually go and use newer, more modern, more powerful tools that are, again, best in class. We can help. We can do a lot of things to help you on your way. One of the things we do, we have a service called a Spec and Select, which is sort of broken down to these five steps here. If you want our advice, I'm happy to give you my advice over twenty minute coffee. I'll buy the coffee, in fact. If you wanna go through a process where you like, we wanna examine the marketplace. We wanna talk to all of our users. We want you to understand what our needs and wants and desires and and shortcomings are, and then make a recommendation. Let's go score it and have presentations and demos and maybe even a proof of value. We can do that too. We're happy to help. There's a whole new future of amazing things that you can do with data and analytics tools, and we certainly want to help you navigate that correctly. The other thing I'll throw out there is we have something that we call a DART. It's called, well, it's a data and analytics review and tactics assessment. And in that we look at five different maturity frameworks across five specific technical domains. That's culture, analytics, data governance, and platforms. And we understand where you are, how ready you are to take on bigger challenges like AI. And based off of some common themes, we'll be able to say, listen, your culture's high, your analytics high, your governance is low. So what that says to me is you've got an innovation culture, but you've got to go back and pay some technical debt or you're at risk. So whenever we see different things, different patterns emerge, we can give you some very specific technical or other tactical advice that'll help get you back on the right track. And again, for anyone that wants that's listening to this, we're happy to give that to you for free, the Dart for free. We love to bundle it as part of any work that we do. So if you're thinking about dashboarding or you're thinking about a data warehouse or the whole migration transformation, we'll normally say, hey. Let's put a Dart in there so that we can baseline everything that we're doing. And at the end of it, let's have another look and see how we did. Look a year later. We're really, really interested in helping. If there's any questions that you guys have or you wanna talk to any of our salespeople, please reach out. We're very happy. Again, we believe in long term relationships. So even if you just wanna spitball some ideas, or Rob, I really like that conversation, or there's some things you said that I don't understand or I don't agree with, happy to come in, have a chat. We've got people all over Australia. So wherever you are, we could certainly come and see you. We've got about five minutes or so. If anyone has any questions, I'm setting myself up for disaster. If someone asks me a hard question or starts making fun of my pictures, I'm happy to answer. All right. We got a question here from ****. ****, it is nice to see you again. The question is if the data doesn't exist in the data platform or in a semantic layer, even in BI three point o, don't we still need to go back to the data engineer business analyst to re engineer things? A hundred percent. A hundred percent. If you think about AI as a wave that's washing up on the shore, think about it laying on the shore in the reverse order of your data life cycle. So the wave is hitting reporting and analytics first, it's going into data preparation and into the semantic layer, then up into the warehouse. I think that the wave is strongly on analytics right now. You can see that with Copilot, you can see that with all the automations in terms of, hey, build me a dashboard that shows me this. That's AI at the analytics layer. I don't think that AI has crested as high as it can as you get to engineering. So one hundred percent you're correct, ****. That's why I said take as many analysts that you can out of the analytics team and put them as engineers. Because the more acceleration you have building curated data sets, the more valuable you're going to have and the more the speed at which you're going to build it. One hundred percent. Very nice comments. So thank all of you for for saying nice things. If anyone has any more questions, otherwise, you can certainly keep giving me compliments. Although my team's gonna have to take me down a few pegs. They don't like it when I get too happy. Otherwise, I might shut it off here. The recording will be up in a couple of days. We will let you know. We will email anyone that registered for the website. So if there's friends, colleagues, whatever that you want to have a view of this, and again, if you wanna have to talk us directly don't be shy but we'll certainly be in touch thanks everybody look forward to seeing you guys again soon bye