Alright. Let's get going. So welcome to our webinar. We are Intoworks. We'll talk a little bit more about who I am and who Intoworks is, but you are here to understand a little bit more about BI three point o and the journey in which it took to get here as well as what's what's coming for folks that are in the analytics world. Before I go any further, there's a couple of little controls I wanna talk about. So, obviously, there's a webinar chat. You can put thoughts and comments in there. I may or may not see them as we go. There's a lot of content, so sometimes I'm looking at slides, but I certainly will try. We'll we will attempt to have, say, five to ten minutes at the end of this to answer questions. So there is a q and a. There is a q and a button. So if you've got some questions, you can put into there. I'm hearing that there is echoing, but I I bet you, saying that that might be on your side. So you might have, you might have an audio issue. Okay. Let's get going. So who am I? Who's this voice that's talking? My name is Robert Curtis. I am the managing director for Interworks over Asia Pacific. I live in Australia. So for me, it is seven AM in the morning over here, and it is a delight be talking to all of my friends in North America. I'm assuming most of you are since we're doing this in a North America time zone. Some of you might be from Europe, some of you might be from other places. Regardless, it is a pleasure to get to see and meet all of you. I've been with Interworks for about twenty years. My, my employee ID, I think, is ID number three. So I've been here for a very long time. I've gotten to, work with a lot of different types of customers and gotten to see a lot of the market and trends as this, wonderful industry of ours keeps changing and evolving. A little bit more about Interworks. We do basically three things. We do strategy. Where do we wanna go? What's our North Star? Why do we wanna go there? We do solutions so we can help build the data governance, AI, analytics solution that you need. And then also, we support. So we support the solution. We can support the applications or or platforms that do that. We can also support communities or individuals in building skills and successful patterns. We are bespoke, we are white glove, we are an end to end solution provider. So if you need any assistance, by all means, please reach out. And we have some incentives at the end of this webinar that hopefully answer some of the questions or solve some of the needs you might have related to your journey to BI three point o. A little bit more about us, and then I promise I'll get into the chunky bits. This is basically an overview of the types of things that we do. This is arranged in a little bit of a data life cycle or maybe project sequence. So, again, we start with strategy, and then we go down to building foundations and data governance and platforms. Within solve problems. Or another way to say this is we add ROI now by applying those platforms and data to solving particular use cases that you might have. We can also help you define what those use cases are. And then when we go down to sustaining and grow, that's us helping you build your community, your data culture, training individuals, and doing application support. We've been in business this year, thirty years. So obviously, that little bullet point needs a little bit of updating. Because we have been around so long and we focus very exclusively on analytics and data, we have a enormous amount of experience and pedigree in this space. So something like seventy five, maybe even seventy seven of the Fortune one hundred are Intoworks customers. In fact, a lot of the vendors you probably are using to purchase software to help you do data and analytics are our customers. A lot of you have probably heard us because of our blog, the Intoworks blog. Whenever I go to conferences or trade shows, everyone's like, oh my gosh. I love the Interworks blog. We get three and a half, maybe even four million page views each year off of a blog that talks about data and analytics. In terms of our experience, we have four thousand customers. We are global. We've got teams here in Australia with me. We've got a whole huge team in, America, United States, and we've also got a whole bunch of folks over in in Europe and several countries. We are not the biggest, but we pride ourselves on being the best. And one of the accolades, we've received a lot of them, was being recognized as a Forbes small giant. So small company that certainly outreaches the number of employees we have and the effect that we're able to do. That's enough of the preamble. Let's get going. So we are gonna talk about BI three point o. And and somehow this this picture of a child is going to relate to that. So the journey from the early days until now. And so you see this sort of reverse progression, this sort of Benjamin Button progression in front of you from old man to young child. But you might be saying, wait a second, Rob. You said BI three point o. Why are there four people from old man, middle aged man, man, and boy? Well, it's trivia time. I'm gonna ask you a question, and we're gonna see how many of you guys can solve this. It's not a hard question, but I want you to tell me what these two things have in common. Chuck it into the chat if you have an idea. So we got this little computer with this application running on it and this little cowpoke here sitting on a little horsey in rural Oklahoma. Any ideas what these two things have in common? We have no one hit to take a gander in our chat. Maybe you can't put in the chat. Ah, there we go. Monica, you nailed it. Both are old pictures. A little hurtful. That's a little hurtful. Monica, you nailed it. Both of these things are from the seventies. So for those millennials, Gen Zed I should say Gen Z. I'm talking to Americans. Gen alphas, and so on. See, useful things did come in the seventies. One of them is me, old Rob over here. I'll be fifty this year. That's that's delightful. And the other one is VisiCalc. VisiCalc is proto BI. It was released in nineteen seventy nine, so it just makes it into the seventies. But this application is important for a couple different reasons. It it pioneered this idea of making decisions using data. It automated a lot of the cool things that people were doing manually for centuries. I think the first book on BI was written, like, in the eighteen fifties in England. But the other thing was really cool about VisiCalc was that this was that killer app that the people that were making PCs needed to justify putting a a PC on every desk in in business. Now, again, for us, seems like, oh, of course, everybody needs a laptop or an iPhone, but they didn't have the applications. They had the technology. It's kinda like AI maybe twelve months ago. We have this amazing technology, but what do we do with it? Well, VisiCalc was one of those first killer apps that justified getting this machine everywhere. So we're talking about Broto BI. We're talking about BI zero point zero. And there are for these older folks on the call like yours truly, some of these names might be familiar. So Lotus Notes, Microsoft Excel, which is still very much mainstay in any sort of analysis, even at the enterprise level, and of course, VisiCalc. There are a lot of things that these early applications did as a pioneer of this capability. So data storage and processing, We've got columns of numbers. Let's add those numbers up with automatic calculations. We can present this data visually. Of course, these were a lot of spreadsheets, but at least there were some ways that they could be organized. Very simple ways to do modeling and forecasting. The application was not so complex that standard business users couldn't get in there and start to do stuff. Of course, you probably did need a little bit more of a financial or accounting background to start to do these things. And you could customize them with very basic formulas, but the path was created. The journey was started. Now there were, of course, horrible limitations. If I were to give you these things today and say, run your business with it, you'd find it quite restrictive compared to what you're used to. Performance was really slow. You basically had, at least with VisiCal, probably just spreadsheets, maybe a bar chart. The advanced analytics did not exist. There was almost no ability to fix or prepare your data. So basically we got off to a start, but there was a huge opportunity for things to improve. Now, when we're thinking about tools like this, from an operating model, all we could really do is sort of this decentralized BI. A couple of people around the office have VisiCalc on their desk. Whenever I have a question, I go stand by their desk and ask, what does this mean? So starting blocks for most organizations. So the thing that's funny is as we sort of go through the different generations of BI, you'll see that we have different options for operating models. But that doesn't mean that these current operating models still don't exist. There are still organizations out there that are decentralized. Some of you might be thinking, that's us. But what that means is is there's really no strategy, meaning people just use the tools as they see fit. They pick the tools, that they like. There's no governance, meaning we we don't have really policy around how we use data, or how we organize the business logic. Overall, enterprise data enterprise data is largely sitting there silent and unused. Sure. We might be throwing it into applications or things like that. But in terms of analysis, no. There's little economy of scale for buying these tools for the enterprise, meaning we get efficiencies and discounts or building the skill across lots of users so that we can sort of build and and and inculcate a data culture. Business logic is capitalized into a few users. So if those people leave or go on holiday, then that business logic goes with them. And what you see in this this BI operating model is there's often just a very few very passionate evangelists which are really driving change. So it's this idea of the lone wolf chasing after the moon, but able to act very agile, because there's really no guardrails around them, but limited in their ability to be effective because, again, it's just a single person. What this looks like organizationally is like this. You have IT doing stuff. And then out inside the business, have individual folks that are just like, I wanna do stuff. I need to do stuff. And I'm a bit progressive in my selection of of technology. And I'm progressive in how I intersect with data. So I'm gonna start doing stuff. Most analytics tools use this approach with free trials to get into organizations even today. So that's BI zero point zero. Born in the 1970s, again, 1970s, beautiful time, produced people like Robert Curtis. Alright. Let's go on. BI one point o. So now we actually have some groundbreaking stuff. I wouldn't say actually BI zero point zero is groundbreaking. We actually have a huge step forward in terms of the ability to put business decision making with data into the hands of business users. So this is things like business objects, Cognos, MicroStrategy, and a whole bunch of other things. But, really, this all started with the arrival of the first data warehouse, which was Teradata. Everything could be put into not everything, but a lot of things could be put into one place. You could prepare that data, again, very simply, and it gave you the ability of starting to do data preparation, slicing and dicing, scheduling predefined, prebuilt reports. There's a semantic layer, an abstraction layer, business logic layer, whatever you wanna call it, a metrics layer inside of Teradata so people could point and have data that's really well prepared and business logic that is defined in advance versus having to try to do it on the fly or just being stuck. When you have tools like this, it is inevitable that IT takes ownership of it, particularly when you've got a data warehouse like Teradata. The ability for data science or the data engineering citizens did not exist yet. But it allowed greater use cases, more complexity. I'm seeing some of the chat here. Okay. Nothing to worry about. And and more advanced use cases that we could go and attempt at solving. Now this wasn't perfect, but it was a huge step forward. Some of the limitations, there wasn't a lot of agility. Do it this way and only this way. The speed of getting this inside out, of course, these are all very complex tools with a lot of neat things that you could do, but the whole user experience was very much still behind. So again, IT had to do a lot of this, which meant there were long queues for getting new reports or developing new datasets. You can only do batch data, only structured data. All the unstructured stuff is invisible still. That's actually a very recent thing. Very high cost because it's the only game in town and high barriers to getting into this. So you had to spend a lot of money and probably approach this in a bit of a waterfall approach to get all of this stuff going. So you were either all in or not. In terms of how this look like from an operating model, Very centralized. Or in in the parlance of this industry, it would be known as a report factory. You have a group of IT folks producing reports. People throw in ideas. They throw in stuff in the queue. The senior executives get their reports first, and, everyone else is probably waiting around. The good news is is that when you have something that's very centralized, it's very easy to put a lot of governance and controls on it. It does make things very slow though to get business insights out to the people that need it. Data development as a result also gets slow because you've got only a small group of folks doing it. The other challenge that you've got is because IT people are doing this. They're not embedded in the business domain, so they don't fully understand finance or risk or marketing or products or supply chain. They understand IT. So they're trying to understand what you want and then building a report. And so oftentimes this is very iterative and takes a long time to understand what they actually are trying to build. And sometimes we never actually ideate close enough to the business to understand. So a lot of opportunities in BI three one point o, but of course, lot of opportunity to improve. This is what this looks like from a schematic or organizational standpoint. So all of the folks up in IT are building your reports, your dashboards, they're preparing the data. So the folks down at the business units are looking upwards and saying, I'd like my report that I requested six months ago. IT tend to be some of the busiest people because everything gets put into that. I have a new, I want to integrate a new application. I want to migrate to something else. The senior executive needs a report because they're over there at a conference. Everything kind of gets pushed onto IT in this model. Now we're into something that looks probably a little bit familiar for most of our folks. We're into this BI two point zero, which is where we've been for maybe the last, say, fifteen years. And a tool like Tableau, groundbreaking, innovative. We're gonna put the business decision making into the hands of the business user. We're gonna give them as many opportunities, and use cases that they can solve. And then you've got tools like Power BI, of course, and Qlik and a whole bunch of others. Massive technology innovations here, not only just from a UI, UX perspective. It's easier for people to learn this. But you can do near real time data. You could do advanced use cases, far more capabilities and predictive and prescriptive, not fully, but more. Visual analytics and interactive dashboard. So I've got parameters. I've got filters. I got dashboard actions. I got all these cool things I can do. It is close to self-service analytics for the business user. The business user probably needs, like, a week of onboarding to really get dangerous, meaning, Tableau Desktop level one, Tableau Desktop level two, and some mentorship and guidance before they go off and start building. Here's our data. Here's how you use our data. Mobile friendly, agile, it can be embedded, and it is just starting to get, AI. So, like, this is your your talk to data, your explain data that you would see in Tableau. It it's there, but not always super contextual and always super useful. Yes. You're identifying some of the records that sit underneath that aggregated data point, but you're not telling me exactly why it's super important. There's a lot of investigation I have to do. Now for as great as all of this sounds, there are still some limitations. Because Tableau and tools like Tableau really focused on dashboard creation, dashboards became the lingua franca, the standardized unit of measure for insights. So a lot of you guys that are on these tools or have been on these tools are probably looking at a catalog of hundreds, if not thousands of dashboards because every time we need a question answered, taching, we get a dashboard. So that means a lot of people build, a lot of people support, a lot of people look at dashboards. So a lot of the organization's weight of analytics people or data people, data workers lean into dashboard creation, dashboard management. But reality is for us to do all this amazing stuff all the way back when we first got a data warehouse site with Teradata, and of course, by this time we have many more options, Redshift, etcetera. The real bottleneck here is, we need more data. Now these tools allow us to do data preparation in Tableau or in Power Query or whatever. But you can see the data starting to seep down into the analytics layer because that's where most people are working. And we're very lightly staffed at the engineering layer. That's not a great thing. But at this BI two point zero, it wasn't a critical flaw yet. It will be in the very next generation. So let's look at the sort of the the most current version of what an operating model can look like. And this is sort of your idea of quote, unquote self-service or federated. All the business units have embedded analytics people, the centralized people, the IT people probably have their own analytics people. There actually might be a separate team that is centralized known as analytics, given, you know, different sophistications of different organizations. The good news here is that business domains could lead their own, analytics. They could even contribute quite strongly to the data strategy. Governance got a little trickier because now it's delegated. Everyone needs to understand how to use data properly. Everyone needs to share definitions. So again, a massive need for data governance tools, but not a lot of payoff for doing it. It became very much a community thing versus something that was very measurable and have very concrete ROI. So a lot of governance tools were bought because we conceptually know we need these, but very few of them became anything more than shelfware. The centralized team still very much owned data. So the data warehouse belongs to IT and all the products that we build out of it, very much IT driven. Analytics went out to the business largely. There was an economy of skills, or an economy of scale. So as everybody learns it, it became this rising tide that floats all boats. But BI two point o and tools like that can be quite expensive. Hundreds of thousands of dollars at enterprise scale. That's probably left there. So what this looks like in terms of the businesses, you've got data workers that are up in IT, of course, that are working on data. So the queue up there is now just building cool data sources. And then we had embedded people down in the business units. I see people saying, this is us. Hopefully we cover off most of you. Now what that meant was is that we have much more useful dashboards because now somebody that understands marketing is building marketing dashboards. So you're able to go a lot further in the depth of analysis and the types of questions you're gonna answer with these dashboards than an IT person could. But there's still queues, and there's still people that are in the gray color there that are data dependent. They need somebody that is a data worker to help them. And, again, this is a limitation of technology. So right now, we've got wave one and wave two and then pre BI. We're here to talk about BI three point o. And some of you probably heard analytics, third wave, BI three point o, whatever. That's what we're gonna talk about now. But just as a little recap, pre BI, proto BI, that's nineteen fifty to nineteen eighty's affected the antecedents, as I mentioned, go all the way back, I think, to, like, eighteen forty with some guy in England that wrote about markets and economies and things like that had graphs that he hand drew in his his little novel. But this is the foundation. It's the explosion of the PC, which then springboard this this CPU processing power and it been doubling every six months. So just this massive march that gave us the technological foundation to do more. And more was wave one, which was the nineties to the two thousands. We could aggregate. We could present big questions or what happened? When did it happen? So we're very historical looking. IT owned a lot of this, highly centralized, highly governed, but we had data warehousing, which was a massive step forward. Wave two, from about the 2010s, maybe mid-2010s to the present. But now this is starting to get a little bit of gray in the hair. The kind of questions were, well, how did this happen? Why did this happen? We could start to ask deeper, more sophisticated questions. There's this idea of self-service for power users, really, but not for everybody. But one of the big breakthroughs here is that the business was able to drive it and we're able to tell better stories because visualization got so much better. Now, before we talk about BI three point zero, it's important to understand how these things are happening. And one of the most important things is cloud native data platform. So we're talking about Snowflake, not a platform that was built on prem that was moved cloud imperfectly, but something that was built natively for cloud. For instance, decoupling compute and storage in terms of the cost, the ability to scale up and down on either of those. Of course, everybody, everybody now knows about ChatGPT and generative AI. Two years ago, nobody knew about it, but now every like my parents know about it and they're in their seventies. But with LLMs came significant steps forward in terms of what business users could do or how technology could support business users. So now I can talk to data. I can do that on Google. I can do that in specialized tools that help me code. This ability to use natural language to get back very sophisticated researched or very technical responses. They're not perfect yet, but they're getting a lot better. The ability for AI to write contextualized responses around data. So it can explain what's happening. A lot of you probably have roles where I take charts or graphs from Power BI or Tableau, and I screenshot it. I put it into a PowerPoint, and then I write about what people are looking at so they know what they're looking at. AI can do that now, and they can do it really well. AI is built into a quote unquote copilot function to help with data engineering, data preparation, data governance, metadata cataloging classification. All of this stuff now is a lot easier and faster, built into tools. The other thing I think is is more than as Tableau took a massive step forward in making I BI easy for business users, the next generation of tools with AI makes it way easier for folks to actually get in and start doing analysis. My, director of sales over here in Australia built a little widget to start to calculate, different returns of different business partnerships. They built an app that you could put in, and they literally just did it by typing it in and did it in about thirty thirty minutes. And and she has literally no coding background. That's the type of stuff that took months to build, can now take days or even just hours. All of these things are happening behind the scenes. So this is the inevitable wave that pushed everything forward to where we are now. Got a question in there. I'll come back around to that saying. Okay. So what is BI three point o? It starts off with your BI tool now natively integrating with all the amazing stuff in the cloud data warehouse. I shouldn't call it a cloud data warehouse. I should call it a cloud data platform because a warehouse or a lake or a lake house is literally the the the foundational beginning steps. They are growing in-depth, but things like Snowflake and Databricks are growing horizontally. They're bringing other things in. Some of the most exciting things are these these AI marketplaces. And when you find a model that you like, you can throw it into a container, and the container can run natively inside your own warehouse next to your data. They're bringing in data governance. They're bringing in data science. They're bringing in all kinds of amazing things. And if you're you're using a BI three point o tool, they integrate natively with that features and functionalities. Of course, AI is augmenting the way that people use these tools. It's not much easier. Say for instance in Power BI, I'd like to see a dashboard about sales last year. Bing, bing, bang. Here's your dashboard. Is it perfect? No. But it's a huge step forward for having to build a lot of this stuff manually. There are more ways of working. What I mean by this is that as much as Tableau, Power BI, and these other tools try to exercise Excel from the from the tool stack, it never worked. Excel is still very much the most used BI tool. So BI three point o tools are bringing the spreadsheet interface and the cell level sort of data manipulation back into a way that is governed. We'll talk more about that a little bit more later. AgenTic AI that can do a lot more. It can actually go and do actions and bring back results. Lower barriers for insights generation because now we have natural language, analytics really thinking of this as a data application interface. So we'll get into more of all of this here in a second. Now, the limitation. AI is highly dependent on curated data. If you have bad data, you're gonna have unpredictable results with your AI. So the the phraseology that I love is that your semantic layer is now the load bearing structure of all of your analytics efforts. Before it was the dashboard, now it's the semantic layer. So until AI is smart enough to translate business logic into a fully articulated semantic layer, it can it can accelerate, but it can't do it yet. Human beings creating business logic as curated datasets are absolutely critical. So the question now becomes, how do we best solve that problem while using all of the benefits of BI three point zero? So the three waves. The third wave is this idea of AI augmented insights. AI basically being your concierge and everyone being a data worker because they've got a dedicated power user in the form of an AI agent or AI accelerated, workflow assisting you. So the we can ask a whole bunch of different types of questions. It's automated. It can be personalized. Again, very bespoke to my needs. If I'm an executive, the AI can be programmed to say, here's some aggregated trends, high level. Whereas if I'm a product manager, it can be it can be, again, trained and programmed to give me far more detail. We can ask better questions. And the process and building into the assets or tools underneath this is all accelerated. So it's a very, very exciting time, particularly for old folks like me that have been doing this for a long, long time. So this is what federated self-service look like. So this is BI two point zero again. And I wanted to bring this back up to sort of contrast what actually things look like, could look like today. And that is everyone is accelerated by AI. The people that are building the data, they can do your your Cortex code where they can type, I need this and boom, Snowflake SQL gets gets produced or Informatica, the IDMC, the new cloud based or DBT. All of these things have cloud accelerators. And if they don't, go use cloud code, go use a specific tool. You can be so much faster. InteWorks, we invested a whole bunch of energy in this over the last two years. And every month, we're finding new efficiencies because new tools, new features, new ideas are percolating up through our ways of thinking. Now, what this does is now you can see down in the business, every single person that was partnered with their power user in the form of AI. It means they don't need to have any more queues. Dashboards, despite what some marketing might tell you, aren't actually going away. They're there to stay. But we don't need ten thousand dashboards. We probably need like a tight fifty because you wanna spend the time and energy to really think through the business metrics and tell a good story. And so a dashboard is very useful. But if I'm a business user and now I have a question or I want a follow-up question, more importantly, I don't need to change a dashboard, create a new dashboard, play around in a sandbox of my own data in Tableau. I can ask a talk to data AI bot the question and the follow-up question. And again, this is all dependent on having a well trained model with great data, But now I don't need a dashboard. I don't need the ability to create a dashboard. I have an AI that's going to help facilitate those questions in real time ad hoc. Very, very exciting. So let's take this and apply it to like a workflow. This is how things worked in two point o. So here we have our data. And you can see if you see a little hand with a click, that's somebody manually doing it. And you see a little green gear, that's the ability to make this automated. Right? So here we have people building business logic. And of course, that's manual. And it probably needs to remain manual even with BI three point zero. We then build a curated dataset. Now that can be automated. We can build orchestration to automate that pipeline and that data shows up in real time or every morning at eight am, whatever it is. And a data engineer oversees those things. And now we build insights in the form of a dashboard again, that used to be very manual. A lot of the work hours I spent over the last twenty years was building a lot of these things. And again, once you live in a dashboard driven work environment, the dashboards themselves become works of art. You can see all these communities. I can't even remember the names of them. They'd be the content. You can make the most beautiful dashboard. That's when you know you are deep in BI two point zero. The analyst, data analyst, BI developer, there's a thousand names for these roles, are building these assets. And the consumer gets the benefit of the insight. Now the challenge is, what if they have another question? Well, oftentimes, that means we gotta go right back to the data and figure out exactly how to solve that problem. And so that might be okay. Maybe we don't have a curated data source for it to replete that whole process. Maybe we need to go and get some of the raw ugly data that data engineers haven't gotten to yet. And if I'm an analyst, well, I have to build all that business logic in my BI tool. If I'm a consumer, well, then again, you guys are probably all familiar with this. I'm downloading it to Excel and doing my own analysis on my desktop, which is even less governed. This is the workflow that I see in all our customers. And in fact, when I did my own analysis for running Asia Pacific, this is a very similar workflow. I needed data, either it was there or it wasn't. And I had to go figure out a way to solve my problem, regardless of whether the tools or data was ready, The needs and urgency of the business doesn't stop. So let me show you what the workflow looks like with AI disruption. So of course, we have our dataset. The data engineer is still really focused on building great data. Again, we can still orchestrate all of that with automation. The transforming and integrating data can be accelerated, but I'm not going to put a little automated there yet because the data engineers using those accelerations to still make human decisions. The analyst, they are creating dashboards, but again, these dashboards are far easier to create. It can be natural language, create me this dashboard, and then they can go do fine tuning. But then we can also build data applications in the dashboard. Conversational AI, contextual AI, which then services the user. So again, if the user has follow-up questions, I don't understand. I wanna see it parsed in a different way. I wanna see a different metric, whatever it might be. If we do this right and we have great data, that question can be sent right back to the AI. Now that doesn't mean that the analyst creates a dashboard and then they're done. They get to go and create a new dashboard. They're still gonna be involved in this, but overwhelmingly, the bulk of the nontechnical report consumer can be serviced by AI. And then the ten percent, twenty percent that's left can go back to the analyst. Let me ask you guys a question. Have you heard of a role, an emerging role over the last, say, five years or so that really came around with the way people were using DBT of the analytics engineer? You've heard of that. Go ahead and chuck some, confirmation into our chat for us. I am going to define it. I'm just seeing if anyone's heard the term. So historically, data engineers to data engineering in a data tool. You're talking DBT, Snowflake, whatever. And then analytics people built analytics in analytics tool. Nice partition, business, IT partnership. With better tools and with a greater need for data, data that is gonna be very centric to the business domains, more and more of your BI people started producing curated data. This is this is a foundational element of data mesh. Business domains contributing to the semantic layers. So it naturally evolved that we're gonna have analytics people that are doing engineering, Hence, the term analytics engineer. If our if our bottleneck is curated datasets and we have BI three point o tools, meaning analytics people don't have to maintain, curate thousands of dashboards. We should be reinvesting their energy, their expertise into engineering more data so that we can do more analytics, machine learning, data science, AI, and so on and so on and so on. So there will be more disruption in terms of not just this workflow, but in terms of how AI bleeds to the entire organization. And it's the smart organizations that adapt that will be able to leverage it the most. I'm working with a lot of different companies on this exact paradigm. So what are some three point o features? Well, let's use the power of the data platform. There's your Snowflakes and your Databricks and do some really cool stuff. You can use real time data versus some refresh that was every morning or if you're really ambitious every hour. Ai throughout the process. AI at the level of solution. AI in the spot of acceleration. All kinds of different ways we can use AI. Companies are building AI directly into the tools for your benefit. We can accelerate data readiness. Again, AI can help do that. You can retask your your BI people to help with that. The tool then now becomes accessible. The other the access to usable data and insights becomes really available to everyone. Everyone can use this regardless of their technical background. You can embed AI and BI into standard ways of working into Slack, into Teams, wherever, you can ask questions directly in the places that you're working versus having to pause and, okay, was this report in the Power BI portal or was it in the Tableau server portal? You can share. It's collaborative. This is because it's got a lot of social media type features where we can comment on it, all these extra things to make this enriched. So let me give you some examples. There's a question in there that I'll answer now. What triggers these disruptions? So if we go back to the slide of emerging trends in terms of the data warehouse, the AI tools, all of these things, that's what's triggering this. It's the capabilities are now present as well as the cloud processing cloud, your AWS, that made such a huge change in terms of, Hey, we've got all this extra processing, we'll sell you the processing. So you don't have to build all the infrastructure. So you can have enterprise platinum level data infrastructure without having to spend platinum level dollars to do it. That was probably a critical thing on top of all the other stuff that we talked about. Alright. So the dashboard. The dashboard's not going anywhere, and it is interactive. It is self-service, but it's it's interactive and self-service at the lowest lowest levels. I got a question on fabric. It's a good one. I'll see if I have time to come back around to that one. So everything that is in pink here is interactive and is dynamic. Right? So you could make all of the other stuff in there dynamic too. But the way that we click and filter and the way we use this is all something that is very interactive. So that's great. That's BI two point zero. It's gonna remain in BI three point zero. But now we have this idea of say a number chart in paragraph or NCP. And that's the section next to it, the AI generated title and all of that text is the contextual stuff that AI is creating so that your senior executor or somebody that's not a data person can read and be like, oh, okay, this is why this is important. This is why that number looks different than last or or this this data point on this line graph is higher than it than I would expect. This is the contextualization, and AI can can start to write that for you so that you don't have to do it. Going even further, you've got this idea of a metric chart and conversion. I'm saying metric versus number because, again, as we're able to enrich that metrics layer, we can have critical indicators, you know, KPIs, that kind of thing that are far more relevant and contextual to our business. Now I, a business user, can ask questions of this data. Okay. That's interesting. What do you think is gonna happen next month? Which product line had influence? Which product line lost us money? Which salesperson is is contributing the best? You can ask these questions directly out of your created dataset and you can build this directly into your dashboard. Because again, these are now native natively accessing the features of your data warehouses. I call this new, but technically, is old workload. So a a lot of these tools, Sigma for instance, is a great proponent of this. You can bring the Excel style workflow back into BI so that it is governed. It is highly performant with millions or tens of millions of records. Whereas anyone that's used Excel, once you get to like your one million mark, everything takes a lot longer. It's prone to crashing. All of the advanced features can be brought in there. So you've got your accelerated formula generation, all those sorts of things. So again, finance as an example, are still that sort of proponent that's like, well, that's great. Your BI tools are adorable. We still do all of our analysis in Excel or Google Sheets. BI is bringing that workload into the tool. Very exciting. Analytics as an interface. And again, I don't know if you guys have ever heard this term swivel analytics. So it's like, you're looking at one screen for your insights. And then when you wanna do something, you turn to another screen. I actually physically just turned here in my chair as if you could see me doing that. But things like I'll give you an example. We built something for Interworks where we had AI looking at all of the different proposals and services that we offer and some of the examples of proposals that we generate. So we were looking at contacts and opportunities, and we said, that's a cool one from the dashboard. That's a cool one. We wanna do data warehousing, data engineering, and maybe some strategy work. Click, click, click, generate a proposal. And boom. From the dashboard, generative AI from our data warehouse produced an example of the proposal. Ninety percent of the way there. We're doing that from the dashboard. That is so amazingly powerful. Generating an email to a contact, whatever it might be, there's a whole bunch of features and actions you can unlock. Generative AI is super powerful in this context. That is a completely new paradigm. Speed to insight. So again, AI accelerating the way that we can develop new data pipelines, data quality, analytics generating new formulas or DAX or dashboards themselves, data preparation easier, faster, more enjoyable, quite frankly. Alright. So let me give you some words of caution, and then I'll tell you guys about some incentives that we're offering. So when you are making tool selections, I think the most important thing if you think about selecting a tool in this context is dating or marriage, you are marrying your data platform. Is yes, there are things like iceberg tables and things like that that make it easy to sort of not easy, but easier to transition from one to another. But it's really a way of working. Snowflake highly managed Databricks, highly customizable. I am not if you so let me let me address the fabric conversation. I'll probably I'll give you my opinion. I am not a certified Microsoft blah blah blah. Fabric has a lot of interesting things in it. It has a lot of interesting ideas. And some of them have only gotten as far as ideas as of today. If you are all in on Microsoft, then by all means use Fabric. If you are looking to put together a tool of of best in class all the way down your data sort of life cycle, Fabric has a lot of promise, but fabric gets very expensive very quickly. So one of the old things that we had to do in BI two point o was talk about the difference between Power BI and Tableau. Tableau was always the tool that was more friendly to business users, which is ironic because a lot of people see Tableau as hard today because of just how much faster these tools have gotten in in accelerating nontechnical people. And Power BI was very much an IT tool. DAX was very difficult. But the big the big selling point for Power BI was this is the tool that IT pays for because it comes as a part of your three sixty five license. Fabric gets very expensive very quickly. So it is on par with, say, Salesforce in terms of cost. It makes sense if you are all in on Tableau or you're or not Tableau, Microsoft. And if you're on Azure. If those things aren't true, then I would say it's probably best to go do a a deep market scan. We can help you do that, by the way. Which again helps me answer this question, this idea of full stack versus best in class. Do want go wall to wall Salesforce? Do you know wall to wall Google? Do you want to pick and choose which ones that are best for you? One of the big things I would say is there are there are legacy tools or BI two point o tools that are saying, hey. We've got AI now. And and and if you want the AI version of our tool, you gotta pay double. That's not that's not realistic. There are a lot of challenger tools that AI comes stock standard as a part of the pricing. They are younger tools. They're smaller companies. Some of these are very small with really, really cool ideas. But AI, you shouldn't be paying a premium for. This is now the expectation. So I wouldn't settle for having to pay a whole bunch more money for this new agentic AI in a very isolated specific version of the tool. And I would also say, listen, you should probably think about, you don't just grab this tool. You really need to re envision the entire way you're working. So we talked, that's one of the reasons I showed you all the operating models. If you're gonna introduce a cloud native data platform accelerated and augmented with AI, and then put BI three point zero on top of it, the definitions of who's doing the work and the types of work they're doing can change. It should change. You should have more BI people doing more data engineering because these tools are now more accessible. And if everything is gonna and run at maximum capacity, you need more curated data. You do not need as many dashboards. And the way that AI has now made governance, data governance essential, it is no longer, we know we need to do it, but it's really hard to point to the ROI on a balance sheet. Well, now it is essential and you can prove that it did. So all of the data governance stuff that's been sitting on the shelf is now out in the field actually earning its key because AI must have it. And I would say this is the thing that I've been saying for years. The most valuable asset you have is your business logic. It is your semantic layer. So if you've got a vendor that is telling you, we will hold your semantic layer for you. In fact, we'll help you build it, but we're gonna build it in a proprietary language. I wouldn't do it. If you can't move your business logic with you as you change tools, then you are trapped. And quite frankly, some of the vendors out there, this is a marketing strategy. So hold on to your business logic as tightly as possible. The good news is Snowflake, Databricks, other things like this. They have sort of this this common language or extensible way to pull your data, iceberg tables, etcetera, as I already mentioned. So that's super important. We can help. There are two different things we're gonna throw your way. And if you are interested, reach out. There are a lot of different ways that you can just interworks dot com on the contact forms is probably the easiest way. If you are interested in looking at your tools, Rob, I am interested in BA three point o. I wanna know how that's gonna affect us. I'd like to get a scan of how our industry is using the current tools. What are the opportunities for us to improve? Or how does this integrate with the other tools that we have? What are the new features out there and help us do procurement so that we can get the best price? We can help you through this. We can do a white glove guided evaluation through any and all of these steps. I just did a five month evaluation for one of our customers down here in Australia. Thirty thousand people. Enormous organization. Because we get to do this every day, it is not new for us. We don't have to go do a market scan discovery. This is our bread and butter. So it is much faster for us to help you than it might be for you guys to learn this all from scratch. The other thing, and we're gonna give this away for free. I don't know if my American colleagues are in here, but, hey, they asked me to do the webinar. So congratulations. You're gonna get this for free. This is this assessment that we created called a DART. And basically, it is a data and analytics review and tactics. What we do is we basically do a bit of a survey evaluation of five critical technical domains. That's data culture analytics, your data itself, your data platforms itself, governance and platforms, and platforms is an infrastructure. We ask surveys. We go across the business and talk to different people that are the technical experts in your business. And then we score them and present it in a visual chart, of the spider web graph, and make observations and tactical recommendations in terms of how you can improve those scores. We also aggregate all of that into an AI readiness score. And really what we're talking about there is how ready are you to to do productionalized AI with predictable success? This takes about a week. So it it is we we do probably need to talk to different parts of your business, maybe five different workshops, but then we come back with a very, very focused, very crunchy, output in in the form of about five, six pages. We're giving that away for free. If you are interested in that, just ping Interworks. You can do that here. Scan that code, contact us today. It'll take you right to the Interworks website, and we are certainly happy to help. Okay. That includes the content. There was a lot of chatting happening while I was going on. So I'm going to go back through. If you do have questions, you can chat them right in that chat, or you could throw them in the Q and A. The Q and A was not as popular today. Everyone was using the chat, which is great. All right. So here we are. Let me just go back through this. I see some observations about data science. Yes. I agree. Here's a question. This is from Sang. Universal skills that go with I don't know if I understand this one. With regard to what's next after data science, deeper context. Saying, I don't know if I understand that question. If you want, Claire, yes, my fiance is an analytics engineer. That's great. Saying, if you want me to answer that maybe, maybe take another crack at writing it. Matthew, at what point in school should our educators begin teaching, or understanding the power of data? Well, should I know they are. Our CEO, a guy named Befar, John Chahi, volunteers his time, I think a couple times every semester. I I think he he even actually had a class. We just volunteered his time to go teach young analytics people the real world of data. So I know there's programs like that all over the place, But I don't know if our educational system is ever gonna be agile enough to capture how fast this industry moves. It's like it's like engineering of any variety. You spend four years in university to understand how engineers think. And then when you get hired, the business has to take eighteen months to two years to teach you their business so that you can actually add value. That's probably a similar life cycle. You understand the concepts of data. You understand the concepts of analysis and data literacy. And then you jump out into the real world and get caught up in the current of the fast moving. You're like, oh my gosh, I'm getting hit by every branch of this force I'm running through until eventually you get caught up to the speed of the race. I see it with the new consultants that we hire all the time. Saying you asked what triggers those types of disruptions. We talked about that. By the way, if you want this deck, happy to share where it will you should, after this webinar, get a email that says, hey, here's the recording. So you'll have the recording of me saying all the words I say, but also if you want the deck, we're happy to share that too. Just looking at some of the other questions. Here's one from Jared. Which tools enterprise or otherwise are currently incorporating the contextual AI you demoed as part of three point o? We have done a lot of evaluation this. And of course, just by definition of our role, we get to see all the tools. We're on industry panels. The one that we are in love with is Sigma. Sigma is a the thing that's cool about Sigma, it does all the things that I sort of mentioned. Spreadsheets, AI generation, talk to data, conversational AI, native to data platform, and it's got parity across BigQuery, Snowflake, Databricks, etcetera. But the thing I like about Sigma is that it is a challenger tool, but it has a maturity level that you can use in the enterprise. Some of these other tools are so small that they really are more appropriate for like a departmental use. But Sigma's got, I think, a thousand employees. They've got hundreds of millions of dollars of funding. They've got thousands of customers. So they are out there proving it in the real world with really large companies. We have fallen in love with them over the last two years and to work. So we are all in. We love it. And we were the very first partner of Tableau, the first global partner of Tableau. So we saw Tableau and the value it had before anybody else did. We were bigger than Tableau when we picked them. We loved this tool. And so we have the same feelings about Sigma. Jared, hopefully that answers that question. Another question, how do you see the balance of exploration via LLM versus persistent assets like dashboards playing out? Will it continue to be a mix of both? Yes. The dashboard is not dead, but the dashboard is overused. The answer is not dashboards forever and always and everywhere. Instead, dashboards are great for your traffic light, your operational reporting, that kind of stuff. But to give people insights on a daily, you don't need a dashboard. You just need the answer. And if an AI bot's able to do that more effectively, and spoiler alert, it is, that's what you should be spending a lot of time. This sort of data app development is is where the next bulk of your energy should go into. Maybe we'll do this one as the last question. This one's from John. Can you comment generally on the work needed to make AI assistants aware of organization specific taxonomies in BI three point o? That is a big question, John. I don't know if I'm gonna have time to do that. I would say that context layer is super important so that you can train your AI in terms of the ways that you need these answers framed, the data that you want it to use it. So it's that whole curation of the model. There is a lot more that could be said in that. So, John, what I would recommend is maybe reaching out to us and we can get you in touch with one of our specialists to sort of go into far more detail. Again, if you just want to spitball some ideas over thirty minutes, forty five minutes over a phone call or a coffee, no problem. If you wanna actually engage for a deeper strategy engagement, days, weeks, whatever, no problem. We're happy to help, however. Claude, what is my what are my thoughts on Claude? We use Claude. We love it. It gives us the ability to build applications to contextually train it on our data and the way we think. So we went fully on Claude. I don't know if I'm allowed to say this. I think within the last few months, we were trying a bunch of different tools or, like, Claude's the one for us. We are at time. Hopefully, this was useful. I had a fun time having a chat with all of you. Again, you'll get the recording you have any needs that we can help with. We are not a pushy sales organization. We have very few salespeople. Most of our people are consultants and technologists like myself. So we are very happy to be, just a advisory, a friendly ear, an idea place for you to sort of ideate without having to worry about buying something. And when you are ready, again, we like to think that we are the best solutions provider out there. Thank you so much. I loved having the chat. If you need anything, reach out. Otherwise, I hope you have a wonderful time and a lovely day. Thanks, everybody.