Video: Precisely Now: Where Data Meets AI | Duration: 3604s | Summary: Precisely Now: Where Data Meets AI | Chapters: Welcome and Introduction (41.035003s), Event Overview (107.22500299999999s), AI and Data Trust (155.395003s), Data Intelligence Platforms (279.475003s), Data Intelligence Context (405.62500300000005s), Data Quality Challenges (521.405003s), Digital Data Steward (719.025003s), Customer AI Journey (943.115s), Unified Platform Architecture (1071.0750030000002s), AI Studio Introduction (1267.635003s), Platform Demo (1438.975s), Prebuilt Agents & Skills (1914.145003s), Live Building Demo (1984.5850030000001s), Closing Remarks (2231.985003s), Customer Experiences Panel (2284.07s), Data Foundation Strategy (2400.015003s), Data Context Importance (2539.425003s), Overcoming Implementation Challenges (2669.245003s), Business Leadership Advice (2957.725003s), Business Partnership Advice (3074.235003s), Closing Remarks (3164.1050029999997s), Closing Remarks (3225.195003s)
Transcript for "Precisely Now: Where Data Meets AI": Welcome to Precisely Now. I'm Matt Waxman, chief product officer, and I'm super excited to welcome you to today's launch event and introduce you to a brand new Precisely. If you're like the data leaders that I speak to every day, your teams have become very accustomed to using their favorite LLM tool on a regular basis. But the gap between using AI for personal productivity and true enterprise agentic workflows is still largely unsolved. That's created a real tension around all of the AI excitement. The technology's ready, but the data it relies on isn't. And AI has a way of exposing that and doing so very, very quickly. Enterprise data is fragmented across systems in the cloud and on premise. It's a mix of structured and unstructured data, and it requires integration, quality, enrichment, and governance to become a trusted asset for AI. Today, Precisely is answering the question every data leader is asking, how do you actually close the gap between AI ambition and data readiness? And this will be an action packed hour with product announcements, with demos, hearing from some great industry experts and some customers as well. We're gonna start with a conversation between our CEO, Waleed, and Stewart Bond from IDC. Stewart recently released the 2026 market scape report for worldwide data intelligence platform software and on the major shifts that are happening in this market, and you can go and download that report right now. After that, we'll show you some great demos of our new products that will help your teams quickly build AI on trusted data. And we'll close with two great customers who are in the thick of this transformation, closing their own data readiness gaps to get AI to scale. So let's dive in. Waleed, over to you. Thank you, Matt. I've spent well over thirty years of my career working in technology companies like Microsoft, Oracle, Ansys, and Sage. And I've seen a lot of transformation around technology and, you know, a lot of technology shifts in the marketplace. All of those transformations would really revolve around getting the technology right, and the technology was the hardest part. But they really didn't deal a lot with the with the essence of what the company is doing in terms of data and data transformation. With AI, this is completely different. It is about technology, and the technology is a very important part of it, but it's also about getting the absolutely trusted data so the technology could operate on. Companies today have massive amount of information about their customers, their processes, their products, their procedures. But a lot of these data data lakes and data islands sit in different systems. They sit in... Some of them are mainframes, some of them are on prem, some of them in the cloud, some of them in SaaS applications. They're all over the place, and they're disconnected. And some of them are stale. Some of them are old. Some of them are not accurate data. And when you unleash AI agentic power on that data, you are not going to get a trusted AI outcome. If you want to get trusted AI, you need trusted data, and that's very important. To that end, I would like to welcome Stuart from IDC who spent many years looking at this problem and dissecting this problem, and he had the research that he published lately talking about this particular topic. Stewart, thank you for being here. You've spent years in this market and the last several months going deep evaluating 15 vendors, talking to buyers and customers, mapping where the category is headed. I want to get your opinion on what's changing, what buyers are getting right and wrong, and where the real opportunity is. Thanks for having me today. I'm really looking forward to this conversation. You're right. We recently did a a... What's called an IDC market scape, looking at the data intelligence software platform market. And for us, data intelligence is things like data cataloging, data lineage, data quality, it's it's it's harvesting all that metadata, it's semantics, it's it's all of the intelligence about the data. It's the thing that you need to be able to have trust in your data. It it... It's one thing to to say the data is no good. It's another thing to say, well, the data has a quality level of 80% or 86%. That gives you a lot more information and intelligence about the level of quality of your data when you go to make decisions based on that data. But I think what's changing, as you alluded to, there... It's significant shift right now, not just in the market, but clearly in in how technology is being used. And and I was at two conferences this week. The common substrate of both of those conferences, because you can't go to a conference today and not hear about AI, but the common substrate was the data. You need clean, trusted data to provide the context that's required for AI. And I think the shift that's really happening is it's around who's consuming the data. It used to be dashboards and analysts looking at those dashboards or working in a spreadsheet. Now it's AI agents that are making autonomous, multi step decisions based on that data, and those agents need to be governed. Those agents need to have governed contextual data at machine speed, not human speed. Right? They can't wait for the data to be ready. The data needs to be ready as soon as that data is available to be used. And that's really a big shift that we're seeing. I would like to ask you a few questions, around this topic. Your new data intelligence market scape opens with an important claim. The data intelligence has gone from checkbox, which we all know, to a first order AI infrastructure decision. What changed? And why is data intelligence suddenly so central to AI? Yeah. So data intelligence is so central to AI because one of the big words that's being used a lot is context. And context is the data about your business. It's the data about the people, places, and things that you are working with, and the things that are really critical to your business being successful. But context is also the intelligence about the data. That data that I'm using with an agent, is it the right data? What is the quality of that data? Where did that data come from? What are the security and privacy classifications around that data? Context, that that intelligence about that data is critical to be a part of the data that's being used with those agents, so you can make sure the right data is being used at the right time, with the right agent, and for the right reason. I'd I'd kind of frame it this way. Data intelligence used to answer the question, where is my data, and is it clean? Now it has to answer, Can an autonomous agent trust this enough to act on it without a human checking it first? That's a much higher bar that needs to be met. It's not a governance nice to have anymore. That's really good point that the the construct of... That the agent needs to make sure that they're operating on a trusted data to allow that so you don't have to have the human, checking in for every decision that the agent makes, and that's how you unlock productivity. Let me ask you another question. So your research found 86% of organizations have increased their focus on data management specifically to support AI. But we all know most AI initiatives are still struggling. In particular, data quality mentioned in your report as one of the widest gaps in the market. Even among vendors that have been in this for years, Why is the data is so hard to get right, and why is it even more important today? Yeah. Absolutely. So data quality has has has always been a problem, and it's never really been solved. And I can tell you that all across many different IDC surveys that are done in different areas of our research organization, potentially even on different primary topics, data quality always surfaces to the top as being a problem, as being a challenge. So I'm gonna take one of those surveys that someone else in Adecee did. It was done by our data storage infrastructure team. I I focus more on the software side, but this was done by the data infrastructure people. They surveyed over a thousand respondents, and those thousand respondents told us, within that population, that AI proof of concepts that have never reached production, 44% of them cited data quality as the primary reason, and another 35% said it played a significant data quality problems, played a significant role. That's over three quarters of the entire population of failed AI products trace the problem back to data quality. And the root cause is structural, it's not for lack of effort. The same survey found siloed data, multiple storage systems, out of date data, and the leading reasons quality breaks down. I... I've I've... And and and those were some of those reasons, but I can also tell you that I've done data quality workshops with clients where it's a process issue or it's a people issue that's called... Causing the problem. It's not a technology issue. It's just... It's very structural on how they do things. I remember doing one workshop where data would come in one certain way, and the person that needed... Was using that data needed to change it, needed to transform it into the way they needed to see it. Only to find out that two steps down the line, someone else needed it in the format it originally came in in as, and so they had to transform it back to what it originally looked like. The problem is that when you're doing those transformations back and forth, things go wrong. Errors enter enter into the solution, and that's where your data quality starts to degrade. It's not tooling gaps, which is why, you know, just grabbing... Buying a point solution and throwing it, thinking that you're gonna solve all the data quality problems with an LLM or with one data quality tool, that's not gonna happen because we've tried that. We've been there. It it comes back to you've gotta take a look at what you're dealing with data, how you're using data, then you can really try to understand what those root causes are. You coined a phrase, which I love, by the way, digital data stored. Can you explain that to to us and tell me more about it? Yeah. Absolutely. So the digital data steward is something where we saw coming from a number of different... Number of the different vendors that we evaluated in this report. And, those fall conference season has just kicked off, I think we're gonna see more. So the Digital Data Steward, it's an agent that curates metadata, it drafts quality rules, detects anomalies, and it recommends or executes remediation. But... And and this is something that the person... The people data steward used to do. But the defining feature here isn't the automation. It's that it exposes its reasoning. It needs to expose its reasoning for human review. But timing without explain... Explainability, it's it's not a data... It's not a digital data steward. It's a black box making changes to your data estate that you might not be able to fix. So I'm gonna bring in IDC's broader AgenTic AI governance research here, which frames this as five non negotiable pillars. The first one, explainability and audibility, critical, you need to understand what the agent did. Identity and access control for the agent itself, the agents are the things, I won't say people, are the things that are working on the data, they are first class citizens in the architecture. Three, a defined human oversight needs to be there. Four, systematic governance across multiple agents, not just one or two agents, but multiple agents. And lastly, continuous monitoring. You need to be able to continuously monitor these things to see what they're doing. And I hate to say you need to monitor an autonomous agent with autonomy, but more likely you're you're not gonna sit there watching what it's doing. You're gonna you're gonna wanna be told about the exceptions when they occur. So a digital data steward has to satisfy all five, and not just log its actions after the fact. For data teams today, this changes their day to day job, it changes what they do. Stewardship is shifting from doing the manual custom work to setting the policy boundaries the agent needs to operate within, and auditing the decisions that the agent made after the fact. The governance question data leaders need to ask their vendors isn't, can it automate this? The question is, can I see why it did what it did, and can I undo it, which is critical? Stuart, thank you very much for doing this session with me and with Precisely. You... You've spent many years going through this area, and your experience has been fantastic. And you've talked to a lot of our customers and, you know, precisely, we're very proud to be placing leaders in this category. Trusted AI needs trusted data. That is the fundamental core of what precisely is all about. We are here to help our customers make sure that they get the best outcome of their AI strategy through the use of our trusted data, platform. Matt, in the next session, will take you through all our offerings and technologies that will make this possible. Matt? Thanks, Waleed and Stuart. I thought that was a great session. For me, one of the key takeaways is that this isn't a theoretical problem anymore. 86% of organizations have reprioritized data management specifically because of AI. And AI projects aren't failing production, not because of the models being wrong, but because the data underneath them actually isn't ready. It used to be that bad data meant a stale report. Now it means an agent takes the wrong action at scale before anyone catches it. So the stakes have changed. For the next few minutes, I wanna share some background on what precisely customers have shared with me and my team about their AI journey and how that's influenced what we've built. One of our primary sources of customer input is our customer advisory board. 20 of the largest customers across financial services, insurance, telco, running data operations at serious scale. When we met in November of last year, the mood was cautious. Organizations were still evaluating, still planning. Our customers told us plainly, we're not ready. We need to slow down on AI. Flash forward six months later, and we met again. Same customers, but that caution was gone. What changed wasn't the technology. What changed was the mandate. AI moved from a strategic conversation into active implementation, and it did it really, really fast. Every single board member had initiatives in flight, and they felt the pressure. The question had shifted from, should we use AI, to why isn't this working yet? And the answer almost universally came back to data, fragmented data they couldn't trust, governance they couldn't enforce at scale, a gap between what their leadership expected AI to deliver and what their data could actually support. We've seen the industry go from talk to action in six months, but action built on data you can't trust doesn't produce results. It actually produces risk. That's the gap we've been building to close. And to understand what we built, it helps to understand the two dynamics that are reshaping this market. The first is a shift from point solutions to platforms. For years, organizations had individual tools, one for data quality, one for governance, one for cataloging, one for enrichment. And analytics was effectively the thread that was stitching it all together. And that worked until AI. Now you're adding AI into every one of those points. And what was complicated before becomes chaotic now. Models and agents pulling from fragmented data in different states of quality with no shared governance underneath it. Unified data management is what brings control to that. One foundation where your AI applications, your agents, and your analytics can all operate from data they can actually trust. The second dynamic I wanna talk about is this architectural shift that's going on. There's a real tension right now between the simplicity of plugging an LLM into a data warehouse and the reality that enterprise data still needs centralized governance. You can't just connect an AI to everything you have and call it trusted. You need a layer that provides context, accountability, and control across all of it. Both dynamics point in the same direction, a unified platform with governance built in from the start. And that's why I'm excited to share the new innovations precisely announced just earlier this week. First, we announced our new precisely platform, one unified experience that brings together data integration, data quality, data governance, location intelligence, master data management, and customer communication capabilities. This isn't a bundle. It's not a repackaging effort. It's a platform that has been built around a single idea. Turning raw data into data that you can trust. And here's what makes it different from other solutions that you've seen out there. The Precisely platform is a control plane, not a data plane. We don't ask you to move your data to us. We meet you where your data already lives, in the cloud, in data lake houses, or on premises, including mainframe, SAP, and all your other established systems. The platform enables you to manage all of it without forcing a migration or a rip and replace. We also built the platform around a unified semantic layer that gives agents and applications the context they need to make better, more deterministic decisions. Think of it as the connective tissue that makes the whole thing coherent. And we have something that pure play data management vendors don't have. We have over 9,000 data attributes across hundreds of proprietary datasets built directly into the platform. Enrichment isn't a separate tool. You have to go and buy and integrate. It's there, and it's available inside your quality pipelines. So you can make your data more complete without adding another point solution. Now a platform is only as useful as what you can do with it, which brings me to the AI Studio, something I'm very excited to share with you all. This is the entry point for the Precisely platform that we also announced this week. AI Studio is where you build. It's where your teams, your AI engineers, your developers can actually create agents, skills, agentic workflows, and AI powered applications using data that the platform ensures is already governed, already trusted, and already enriched. Think back to Waleed and Stewart's discussion of what it actually takes to get AI into production. The challenge isn't building a demo. The challenge is building something you can trust at scale inside of a complex enterprise. AI Studio is built for exactly that. It dramatically lowers the barrier to building AI applications on top of governed data. We prepopulate it with AI agents, apps, skills, and more. And over time, customers and partners will contribute to it as well, making it a community of builders sharing tools with one another. And it's all MCP native, so it connects to the model providers that you choose, the tools, and external apps that your teams are already working with, not as an afterthought, but as an actual design principle. The AI Studio is available today to help you quickly build solutions for your data challenges, and it works with either the existing data integrity suite or the new Precisely platform, which is now available in preview. Before I send you to the demos, there's one more piece of this that I wanna mention. The Precisely platform isn't just infrastructure. It's a foundation you can build real solutions on, and we're gonna illustrate that ourselves. One of the first workflows we're building on top of the platform is multi domain master data management. MDM, as it's known, has traditionally been one of the most manual expensive data disciplines to maintain. And we think that AI changes that, and the platform is exactly the right place to build it. You'll get an early look at the MDM workflow in the demo today. And this fall, we'll be working with a select group of design partners to explore it further. So the Precisely platform, Precisely AI Studio, and a glimpse of what's possible when you start building on top of it. That's what we're sharing with you today, and we're very excited about it. Your AI is only as good as your data. That's more true now than ever. And with agents beginning to act autonomously, the organizations winning with AI aren't the ones with the most models. They're the ones who build a trusted data foundation first. And that's what we at Precisely are here to help you do. And now, I wanna show you what it actually looks like. So please join me in welcoming Fernanda Tavares and Megan Hines from our engineering team who are gonna show you some awesome demos. This is the part of the event I've been most looking forward to, the opportunity to sit down with Megan and Fernanda from our engineering team and actually see these new products in action. So thank you so much for joining. So we talked a lot in the earlier parts about this journey towards creating trusted data for AI. And, Fernanda, I've been super impressed with how all of these products have come together in such a short period of time. So maybe you can give us a little bit of insight into the factory. Like, how did we actually build these things so fast? Yeah. You're right. Earlier this year, when we were thinking about building something so transformative and putting precisely platform together in such a short amount of time, it sounded insane. Right? But because we were using AI, we were able to make it happen. I love it. I love it. So, Megan, I think you're gonna take us through the brand new Precisely platform. Yep. So over to you. Alright. Well, thank you guys both so much for having me here today. I'm really excited to be sitting down with you both and talking more about the Precisely platform. So as Matt shared, everyone is moving fast with AI and agentic initiatives. But the issue is trust and data is still really low and governance that's necessary, to be successful just isn't existent. So today's demo isn't really just gonna be a lot of demo fluff. It's really gonna be about proving that your data can be made AI ready. And there's two parts to that. So the first is we wanna find out is your data good and then can you prove that. So let's start with that first part here. As a leader on my team, whenever I need to understand something more deeply, I'm gonna look at the data. Now the problem being, I don't work on our data operations team. So for me, whenever I'm trying to get an answer to a question, I kind of have to go start pestering one of the members of our data operations team. You know, what data do we have? What do I have access to? Where is it? And then how do I get to it? And with the precisely platform, that's all gonna be in one place. You can see what you have, what it means, where it came from, and whether or not you can actually use it for the purpose that you're trying to use it for. But finding your data isn't enough. We also wanna look at what good means. Is your data valid? Is it complete? Is it accurate? And we also wanna watch over time for data drift and freshness of your data. So when we go in here, every asset you see is gonna have a quality score. Every asset also gets a governance score. It's not just a document, but it's also describing intent of the data. Say I can finally get the data, but that's still not the end of it. Right? Because I found it isn't going to be the same as I trust the data. So here's the second part. Can we actually prove that this data is good? And so every asset in here is going to carry a score, and it's not gonna be someone's opinion in a Slack message. It's gonna be an actual number of your data. So you can have a quality score, a data governance score. And when I ask a question based off of the data, I'm not just gonna be blindly trusting it. I'm seeing that score and how we got to it. And I see a lot of datasets here, Megan. Where are they coming from? Yeah. So here's where it's really cool. This isn't just gonna be your nice clean cloud data. This is gonna hold things that most platforms can't even touch, your mainframe, your IBMI. And that is really, really important because 92% of enterprises are keeping data exactly where it is and that's on purpose. It's not technical debt, but a strategy. So everything that's running on this data runs where the data actually lives, not just wherever is easiest. I love that, what you showed us, Megan, really exemplifies creating an integrated experience inside this platform that's really oriented towards a customer outcome. And one of the things that we're actually building on the platform is a multi domain master data management workflow. It's gonna help us answer questions like whose customer record is right, whose product definition is right, and it's gonna take those definitions and flow them straight into the ERP and CRM systems that you already run on today. We're gonna be working with customers to continue to refine how this workflow will function at precisely. So, Matt, something I'm really excited about is that the next time I need to answer a question with data, I don't have to go bug someone on our data operations team. I'm already gonna know if it's good, and I can already prove that it's good. I'm sure they're really gonna appreciate that as well. So thank you for giving us the sneak peek into, what's to come. It's a great example of an app that we can build on the platform. So I wanna switch gears a little bit and talk about our AI studio. We heard a little bit earlier about this opportunity for builders. So, Fernanda, again, I think the perspective from the engineering team is really interesting here because there's a lot of parallels from what your team's gone through and what customers are going through as well. So how did this idea come to be? Precisely, we build tools that get your data ready for AI, but we are also just a team that builds with AI every day. So a lot of this came out of watching our own engineers work. What I saw on the team was that a group of people naturally fell into being builders and started building agents on their own while another group was struggling to get started. So AI studio is for those people looking for a quick way to start. Agents, apps, and skills you can run on real data in minutes, and it's running on the platform that you just saw. So I do mean minutes, which we are about to prove, Matt. I love it. So, Megan Yeah. Please do us the honors. Show us AI Studio. Well, I am so honored right now to be showing you guys this. So right now, we're on the AI Studio landing page. And the easiest way to think about AI Studio is sort of like recipes. It's a precisely cookbook. We have easy ways to get up and running with precisely capabilities without having to learn it inside out first. We have agents sample applications that people have built on top of precisely these APIs and skills that call the platform MCP for specific jobs. It's organized by industry and data domain so that you're starting from a problem you're actually trying to solve for. AI Studio gives builders examples of applications, agents, and skills that are built on top of precisely capabilities, whether it's a property analyzer, fleet management application, or an industry specific workflow. These examples show how you can use trusted data and precisely services as building blocks for your own AI solutions. These aren't meant to be boundaries, but these are more like starting points and inspirations for what you can build. So one of the things in here is sample applications. Everything in here is built on our MCP. It's the same platform underneath, which is really the point of the AI studio. So Precisely has been doing location and data management for decades. That part is ours, and it's really, really hard to replicate. But what's new is that you don't have to consume it our way anymore. You can use the platform and the UI that we showed you, or you can take that same capability, and you can build whatever you want on top of it. So this app is one person's idea of what to do with that. Yours might be different. So it's our expertise and your creativity. So can we see an example, Megan? Yeah. So this one is gonna be a full property analyzer, and it's gonna have a property report for something like an acquisition or a lending decision. You can use it to look at portfolio health and risk. You can have on demand analysis for any US address. So I'm gonna go ahead and run an address here. Actually, the Precisely office 1700 District Ave. And so this is all live straight off of the MCP. Think of all the really cool possibilities here. Megan, this, to me shows so much of what we do to help customers solve problems from doing it in a very agentic friendly way at the same time as using the recipe book. I love your analogy around that and solving real problems, like actually finding out information around a property around an address. Super cool. Yeah. No. Thanks. And so that... That's just one of the things in here. There are two other things that I wanna show you both today. So not everything needs a whole application. Sometimes you just wanna get a job done and you can use an agent to do that. So we have prebuilt agents that already work with our MCP, like that app. And like the app, they're also examples. So this is what an agent built on our MCP can look like. So you can take one as a starting point and then build on your own. So run what's here or use it as a pattern. And then the last thing that we're gonna have here are skills. So skills are just gonna be instructions that teach a model how to use our MCP for a specific job. These are out of the box, so you're not writing your own from scratch to go ahead and get moving. Grab one and you can use it right away with the MCP. Or read through it, and it can give you an idea of what you might actually need. And it goes into your own setup. So wherever your team already works. You can build this once on trusted data and then run it where you want. So that's really it for AI Studio, but you shouldn't just take my word for it because I've done this before. But, Matt, I don't think you have, have you? I have not. Are you springing something on me? Maybe. Are you okay with that? I'm totally fine with it. Yes. I'm a builder at heart. There we go. Alright. So, Matt, I'm gonna hand it off to you and get you building in here. Very good. So, Megan, you're gonna have to give me a little bit of instruction as we do this. Yeah. But, I suppose I start by going to the AI studio? That would be a good place to go. Alright. So we're gonna start by signing up for a trial here. That's super simple if you wanna sign up for a free trial too. Where do I start? So first thing you're gonna wanna do is connect. Right? So we wanna connect, the Precisely MCP to wherever you wanna use it. Okay. I'm I'm a Claude user. Is that a good place to start? Yeah. Yeah. Okay. So I see here that in the connection dialogue, I can connect directly with Claude desktop. I just need to generate a key around that. I suppose that's the step to do. Yeah. Do that, and then hit that button that says download for Claude desktop. Okay. Let's, get that going here. I've got my quad desktop, loaded. So where do I go in here? I suppose I go into settings to start to load in an extension? Yep. There you go. Great. You can tell I am, I guess, a bit of a builder at heart. I have a little bit of instinct around this. So I will get the Precisely MCP server loaded. This is what's gonna get me connected to the Precisely platform. Right? Yes. Exactly. Great. And it's gonna ask for my API key. Of course, we're not gonna share that live on this event, but I will, get that copied over and put in the, secret. And is that it? Am I just connected at this point? You're gonna wanna go ahead and just hit that enable toggle. Okay. Great. All set? Yeah. Let's double check that. So x out, and then we can go ahead and just double check that that's working by looking at all extensions. Yep. There we go. It looks like it's loaded and configured. Yeah. Awesome. So you showed a property analyzer, example. Yeah. So can I load the same thing Yeah? Or a skill around them? Maybe it's an... A simpler place to start? Yeah. Whatever you want. So, yeah, you're on the property analyzer skill right there. Go ahead and jump into that card. Got it. I've got the property intelligence skill. Cool. I see that there is, a markdown file that explains how this skill works. Yeah. Just to get an overview of it, and you can also just download the whole thing. Right? So it's easier to get...... Easier to get loaded right into Claude. I can just download this directly. Yeah. Very good. I have done skills before, so I'll admit on the builder side of things. So you you keep me honest in here, but I think I can just upload a new skill, into Claude using that property intelligence, right out of the AI studio. Yeah. Exactly. Looks like I've got that installed. So how do I give it a a run here? Let's test it out. Should I do a property intelligence skill? Yeah. And we're gonna check the address. You were using our HQ. Right? I was. I was. Okay. I happen to know that because that's where we are today. So I'm gonna ask for some information around, this address. And tell us what's happening behind the scenes, the... What's the magic that's going on here? Yeah. So we're calling the MCP the MCP that the platform is based off of. So this is all kind of like the same thing, really. And now we're gonna look at, what actions there are in the MCP, and then we're gonna call that skill to format it in the way you want it. Very cool. Great example of using natural language with a governed platform with data enrichment capabilities to actually pull it all into a very conversational experience. Right? Yeah. It looks like it's coming back with an answer right here telling us, a property report on this very building and address Yeah. That has some very interesting information, including tax information in there. Our, property attributes are endless. Yes. This is great. So, how did I do? I think you did okay. Fernanda? Yes. So what if I cannot use Claude or if I'm on Teams or Slack, can I use RMCP with something else? Yeah. I love it. So this is a great example, I think, of... For the builder experience, like you said, Fernanda, for all of those that are looking to get started, or early in their journey building agentic workflows, prompts, skills, so forth. Really exciting stuff. Definitely appreciate you both, joining to show us, all the great new product capabilities that we have. And, for all of you, please go try this out. This is live on the Precisely website. You can log in, use the AI studio, and use it the same way we just did today. Awesome. Thanks so much for giving it a try, and thanks for having us. And we can't wait to have feedback from other people trying it the same way that you did. Sounds great. So now you've heard about the rapid shifts in the market being driven by AI, and you've seen what Precisely has built. But rather than hear more from us, I want you to hear from the data leaders who are doing this work firsthand. When you try to apply technology inside a business that already has years of systems and processes, data and expectations behind it, that's where the interesting conversation really happens. And every organization is navigating their AI journey differently. So next up, you'll hear from two members of our customer advisory board. Kevin Guillot, vice president of data analytics at Lake Michigan Credit Union, and Janine Roberts, vice president of customer operations at Charter Communications. They'll be joined by Dave Woods, our own SVP of global services. And together, they'll have a candid conversation about experiences and lessons learned as they navigate data readiness for AI in their organization. So, Dave, Kevin, Janine, over to you. Thanks, Matt. So, Kevin and Janine, first, I'm so glad you're both here. I really wanna be straight with the audience about why we asked you two specifically to join us. So I've been in the room with your teams when things got hard. When the AI mandate came down really fast and the data wasn't where it needed to be, instead of spinning it, you both leaned in and figured it out. And that's less common than you think. So what Matt just showed you is what we're building. What Kevin and Janine are gonna tell you was what it actually looks like on the ground. The real decisions, real trade offs, the moments where you have to stop and fix the foundation before you can build anything meaningful on top of it. So now with that, let's get into it with Kevin and Janine. So, Kevin, I wanna start with you. So we've talked a lot today about how fast the AI mandate has shifted, right, from strategy to active implementation. So what did that shift look like inside LMCU, and what did it ultimately expose about your data? Thanks, Dave. Yeah. The real shift wasn't, was more about who could use AI. It went from a specialized tool that we have, for data scientists to leverage and use to something really anybody could use throughout the organization and as simple as a chat interface. And overnight, AI wasn't really this project on our road map for us any longer, but it was really in the hands of every user. And that changed everything. We weren't debating about what type of models we were using any longer, but we had to prove and continue to prove that the underlying data was trusted and consistently defined and ready for real decisions. AI started to really expose how poor our data quality was and our data definitions were across the organization, faster than it ever had before. And teams that were continuing to do this manually just weren't keeping pace with the speed at which the organization was trying to move. And so our focus really started to shift from doing these AI projects one at a time to really focusing on providing a governed data foundation where we're sharing definitions across the organization. We're creating certified datasets and data products and semantic models so that data could be easily accessed, not just the access, but trusted, and trusted so that the business had data ready for them to use. Awesome. And Janine, how about you? We're at a different, point of the journey than than Kevin. We follow a very structured approach over the years. We've been playing in the AI data space for a while. We have an agentic agent that sits in front of customer calls as an example. We have an agentic that sits in front of, chats with with customers. But underlying data was critical for our strategy foundation. And so we've spent a great deal of time and investment at understanding our data and cleaning the data. And now we are at the place where it's... Alright. Let's hurry up and go, and how do we move from having all of this now structured, cleaned, governed data available to how do we implement AI strategies beyond what we've already put out in front of our customer base? How do we execute internally on... And process improvements on finding efficiencies and really just changing how we do business. Awesome. That's terrific. Yeah. And, you know, it's... I think that, you know, what I found is, you know, always having good trustworthy accessible data has been important, but one of the changes from an AI perspective, especially with the GenicAI, right, where, you know, some of these models are driving decisions autonomously and, you know, ultimately authoritatively is context. Right? And, Kevin, you know, you you also talked about kind of making the, you know, common definitions across the business, common understanding. But is context also important and part of that journey for you guys as well? Yeah. I'll start there. The context is absolutely really important. Without context, even ourselves struggle, and AI only exacerbates the problem, because of how fast it moves. And so if you ask anyone across the organization for the context around a metric or KPI that they're looking at and the data that backs that, without that context, it's hard for them to understand that metric or KPI and how to measure towards that and how to actually drive the underlying processes that need to be improved to help achieve those metrics and KPI. And if we can't write that down and document that and provide that to these AI models or AI chatbots, it doesn't stand a chance just like we don't of providing good quality information and answers back to the business. Yeah. And Janine, thoughts on that? I know you've always felt like that was, you know, kind of the missing link was the context of the data. Yeah. Huge huge, foundational element is context. As Kevin state... Stated, it's not just AI. Right? It's just in how we should be doing business. And so think about all the time somebody outside of your immediate work work function has gotten their hands on a report about your function, and they are interpreting what they are seeing without context. Sometimes that leads to long decisions being made and a whole lot of work for somebody to try to go and clean it up. So we can't... We cannot emphasize enough the importance of defining context of the data. Otherwise, poor decisions, bad answers are going to get out there. That's great points for both you guys. So let's shift a little bit. So, you know, every organization hits a wall at some point in this journey. So what was your wall that you hit, and how did you work through it, and what lessons did you learn from it? Ginny, maybe I'll start with you. Sure. I I don't know if we're at a a wall yet. There could be a wall that we've hit for a for a particular function or particular model that we were rolling out, but in in totality, we have not hit that wall yet. How do we get past the moments of I don't think I can go any further? We we test the data. We test the answers. We test the models. Is it providing what we are expecting? Where are the flaws? It's it's not unlike other process improvement efforts where you have to reevaluate the journey and make sure that you have all the right pieces in place to get the outcome that you're hoping to deliver. And so that's been a a strategy that we always deploy. Right? Look, measure, learn, test Yep. Buys. Yep. Yeah. That's great. And and I think you bring up a good point. Maybe it's not a wall. Maybe I didn't frame that correctly. I think of it as more a little bit of, like, white water moment. Right? So, you know, as everybody start this experimentation with AI and everybody's kind of using some of the new tools and, you know, we're having some good results, right, that are meaningful results or fast results. And then to scale that into production. Right? And then to deliver business values and AI solutions that are part of your business strategies, you come to this chop part. Right? This whitewater part. And how do you navigate through the whitewater, I guess? So, Kevin, kinda with that reframing, you know, what are your thoughts for all MCU? Yeah. I think for us and for many organizations that I've been a part of, it comes down to being self self aware and having some sort of, I'll say, emotional intelligence around really reflecting about the organization and company and about how your processes and and how the people define those processes. And what I mean by that is we've got access to technology. Technology is everywhere, and implementing that technology in one organization or other may be easy, may be hard, but for us, it wasn't wasn't difficult. But but the hard part was that self awareness piece about everybody actually sitting in a room and coming to the realization that the way they define a customer or product or for us, maybe a member or a service may not be the same. And if we don't have that awareness and the courage to speak up and say, look, if we can't come to the same page and same terms, it's gonna be hard for us to be able to to move forward together. And so I I think for us and our team, it was spending the time to work with the right level in the organization at a very senior leadership level, pushing down that ownership to them and down through their organization on what those definitions were and and how we drove that. And we took all of that information. We really drove that through all the different data models and data products or semantic models and the context that are provided, like, back into the organization into their hands and really drove ownership for them. It comes down to meet people and and the process around it more than it ever comes down to the technology. And the organization, at least many I'm in, don't really understand the governance processes really well, and so it was a ton of education. We spent more time educating and bringing that self awareness. Call it data literacy. Call it data fluency. Call it what you will, but it was education around what this means and who really owns it. Quite often, we find that they think IT or the data team owns the definitions and data, but it's not... We're not the ones helping to drive and grow that organization on the front lines. We're supporting them behind the scenes to really help to drive that organization. Yeah. Just just to add. Right? So Kevin and I sit in different spaces in our organizations. I I represent the business. And so, defining how we will use, what the data means, the use cases is part of what I'm responsible for. Working in partnership with my IT support team that can help me then govern the data model, the data sources of truth. And I have responsibility for the company to understand who needs access and help educate, as as Kevin stated, those people on the purpose, the meaning, the why behind the data that we've now made available to others. Sorry, Kevin. I didn't mean to step on you. I thought I would just want to add. I I think there's a perfect ad. Right? And you're just taking the point home of when it's done well, which Janine's doing an amazing job in her organization driving at, real value gets pulled out of it. Yep. Yeah. I love kind of, you know, the... You know, first, I think, you know, as it relates to AI and and obviously both... Knowing both of you, kind of the the lines between traditional IT and business are a little bit blurred. They continue to get a little bit more blurred, I think, which is a positive thing. But also, you know, you know, provide some opportunities and some challenges as well. So, you know, kind of with that, Janine, I'll stay with you. Right? From a business lens. Right? So a business leader that has accountability responsibilities for, you know, data integrity in their data programs or the outcomes of good data, you know, on their business process execution layer. Now what advice would you give to a data leader, you know, with kind of a business lens as they embark on this journey or, you know, they try to scale, you may... And, you know, kinda up level from, you know, what's probably some early innovation and prototyping around AI? I I think some of the fundamental things that we've spoken about today, this isn't very different from other business process improvement efforts that you may have embarked at some point in your career. Hold on to those truths that you learned a long time ago. It is important to have a strong foundation, and for an AI case, that's gotta be the data. So how do you build the data that's strong? You test it. You govern it. You retest it. You ensure compliance across your ecosystem. Chunk this out into small steps that are manageable, be clear, and be transparent as you go through this journey. Involve the key business partners, technology partners, and data partners in the journey. Don't be surprising somebody at the end. Sit... Tell the... Your business partners what your North Star vision is so that they can understand where you wanna head Yeah. But also chunk it so that you can deliver in in meaningful, manageable steps that deliver value for the organization. That's a great... I mean, that's a great approach. You know, so, Kevin, how about you? And I know you've been kinda on both sides of the house and you've been through these journeys. We've known each other a long time, whether it was, you know, data lakes and lake houses, and now, certainly, the pace and cadence of this AI transformation. But I think it's a lot of commonality and lessons learned as Janine pointed out as well. So how about you? Like, if you were gonna sit down, I know you're pretty sought after and you talk to a lot of peer organizations. You know, what do you tell data leaders that, you know, are not quite far as long on this journey as you? I think Janine provided a lot of good guidance. And most of what she said, I would absolutely reiterate. And we can get caught up really easily in these emerging technologies and how cool they are, and and we can focus on possibly what it could solve and could do, but we need to get close to our business. We need to partner with them and really understand their challenges and problems and help them solve those hard problems. The environment's never gonna be perfect. The data's never gonna be perfect. And sometimes waiting for it to be means we miss that opportunity. But if we partner with the business and we know their challenges and problems and we level set that upfront, we still can drive really great value for the business and help them grow. And so for me, that's that's my advice to the leader. Don't just stay inside your comfortable zone of IT or technology or data. Get out there with the business. Understand the real problems that your organization has and partner with them to solve for them. Yeah. That's great. So I gotta say... So, Kevin and Janine, I mean, thank you. Like, this is the kind of conversation that this industry just doesn't get enough of. Right? I've been doing this for a long time, and the pattern I see is consistent every time. Right? Organizations that win with AI aren't necessarily the ones that are moving the fastest. They're the ones that had the discipline to get their data right first. You know, I was out, you know, last week meeting with a a senior leadership team and, you know, one of the global supply chain leaders, you know, had a really great comment that really struck me as part of this conversation. He said, hey. You know, I'd rather have no GPS than a GPS with bad information because all it's gonna do me is... All it's gonna do is get me to the wrong place faster and with more authority. So, you know, I think, you know, you guys understand that, you know, you're gonna pay for those decisions later if you don't get the data right. So, you know, I'm really, really fortunate to have you guys, spending time with, you know, with us and our team. You know, your teams understand that. You guys are great leaders in that regard, you know, and that's quite frankly why you're as far as long as you are. So with that, I'll hand it back back to Matt. Thanks very much. We covered a lot of ground today. We talked about where the market's heading. We saw the Precisely platform and AI studio that we just announced this week, and we heard directly from data leaders that are doing this work in the real world. What makes me optimistic is that the path forward is becoming more clear. Organizations succeeding with AI aren't necessarily the ones with the most models or the biggest pilots. They're the ones investing in the foundations. They're making their data more accessible, more reliable, more accountable, and more complete. They're building the confidence that allows AI to move from interesting discussions to meaningful action and business outcomes. But I wanna leave you with one thought. If there's a common thread running through all of these conversations, I think it's this. The barrier to AI was never gonna be the AI. It's the data. For years, organizations could work around fragmented data. People filled in the gaps. Teams reconciled the inconsistencies, and subject matter experts provided the context that systems needed. AI changes that equation. Agents don't stop to question the data they're given. They act on it at scale, and they do it far faster than people can intervene. And that's where Precisely is focused, ensuring that you have a solid foundation and a fast path to building AI applications on that data. Thank you for your time today. I encourage you to head over to our website, try out the AI studio, watch our demos, and reach out to talk with our team. We're always excited to learn more about your goals with AI and how we can help you achieve them. Them. Thanks again, and have a great day.