Video: End Shadow Analytics and Regain Control Over Enterprise Data With Dataiku | Duration: 2328s | Summary: End Shadow Analytics and Regain Control Over Enterprise Data With Dataiku | Chapters: Welcome and Introduction (6s), Dataiku's Four Pillars (132.375s), Dataiku Platform Architecture (308.26s), Governance and Collaboration (481.17s), Live Dataiku Demo (668.62s), Shadow Analytics Prevalence (1847.085s), Data Governance Measures (1922.41s), Preventing Shadow Analytics (2079.4s), Conclusion and Thanks (2234.42s)
Transcript for "End Shadow Analytics and Regain Control Over Enterprise Data With Dataiku": Okay. Well, hello everyone and welcome. We're so excited to have you have you here with us today for what's shaping up to be a critical conversation for IT leaders everywhere. As we wait for a minute or two for everyone to fill in, feel free to say hello on the chat. Maybe let us know where you're joining in from and also what role you play in your organization's analytics efforts. Before we dive in, just a couple of housekeeping items. If you have comments, feel free to share them in the chat. If you have any questions, drop them into the Q and A section. We'll save some time at the end to answer as many questions as we can, and we'll try to answer them as well, along the way. And yes, this session is being recorded, and we'll send the recording and slides to your inbox after this webinar. So let's talk about shadow analytics. Whether you call it shadow IT or decentralized decision making, the result is the same, and that is a lack of visibility, control, and consistency across enterprise data. Without governance, analytics workflows can come can quickly become a black box. And for IT teams responsible for security, compliance, and scale, that's a serious risk. And so today's session is all about helping IT take back control, without slowing analytics teams down. We'll show you how Dataiku, the universal AI platform, eliminate shadow analytics by providing full visibility into who's using what data, where, and how. Second is how it provides a governance first alternative to desktop based platforms like Alteryx, ensuring stronger security and compliance, and also how Dataiku enhances analytics with data AI powered features for faster, more reliable insights. Next slide, please. We will also have our wonderful Cody Brooks, solutions engineer here at Dataiku today, who'll walk us through how this works in practice, including a live demo. So with that, let's get started. Cody, I will hand it over to you. Awesome. Thank you, Renata. Hopefully, everyone can hear me okay. Like Renata said, today we're gonna be going over why Dataiku is here to put an end to shadow analytics. But first, let's talk about how we got here. Right? So, typical customer journey, starts when organizations begin to realize they need, to change the way they do analytics. This often begins, with teams adopting tools like Alteryx for self-service data angling, spreadsheet automation, and really quick easy wins. It's really easy to get started, and onboarding feels really seamless. As adoption grows, challenges emerge. What started as a simple solution turns into fragmented ecosystem. You have data duplication and proliferation. Workflows remain constrained to desktop environments, really limit in collaboration, and IT leaders really struggle with governance, visibility, and access to Cloud resources. This is really where Dataiku changes the game. Instead of a collection of isolated workflows, Dataiku is going to provide a centralized, scalable, and governed platform, ensuring IT and business teams work together seamlessly. What is Dataiku? If you're here, we'd like to really break it down into four major pillars. Dataiku really is here to unite those four critical pillars to ensure, AI and analytics success. The very first pillar is your people. AI analytics isn't just for data scientists, isn't just for data analysts. We're here to empower everyone from business analysts to engineers really by offering a platform that accommodates both code and no code users. Whether you're a builder, consumer, coder, or a clicker, Dataiku provides a single workspace for cross functional collaboration, which means faster insights, fewer bottlenecks, and a more AI driven culture. Next is going to be your data. AI is only as strong as your data that fuels it. With Dataiku, you're not locked into a single storage or compute environment. You can work with structured data, unstructured data, on prem, in the Cloud, across multiple engines, really all in a way that ensures performance and scalability. Next is gonna be your governance. As AI adoption grows, so do concerns around security, compliance, and risk. Dataiku is providing you a built in governance across the entire data to AI life cycle, really ensuring visibility and control from data prep to model deployment. Last but not least, your technology optionality. No one ever wants vendor lock in. Right? So Dataiku is really built for flexibility, giving you the freedom to choose what LLMs, what data platforms, and really any cloud provider that work for your business, both now and in the future. Whether you're integrating open source tools, experimenting with new AI frameworks, or standardizing enterprise solutions, Dataiku meets you where you are. So where does DataIQ kind of land in the, ecosystem of your organization? Once again, you know, any data anywhere, no matter where your dataset is, DataIQ is gonna connect seamlessly to it. Really where we see Dataiku is it being the end to end AI workflow all in one platform. I really like to put your eyes right to the center of the screen right here. Didn't mean to do that. Sorry. You can see there's two main points to where Dataiku lands as a platform. First, starting with the design and experimentation. Low code, full code, all the users coming together to build AI solutions. Data engineers, scientists, and analysts can handle data prep, analytics, machine learning, Gen AI, all in a single place. Next big thing that comes with the, Dataiku platform is gonna be production and orchestration. So deploying AI projects shouldn't be a bottleneck. With built in CIC monitoring and governance, IT leaders can ensure AI models are deployed sufficiently and responsibly without operational chaos. We really want to make sure that Dataiku is, has a seamless integration with you and your business apps too. AI isn't just about models, it's about value. Dataiku integrates with third party BI tools, APIs, and your own business applications. Insights move seamlessly from experimentation to decision making. How does Dataiku, how do the nodes work together? You'll see all of this in the demo here in a little bit, but let's talk about how Dataiku's architecture enables seamless collaboration, automation, and scalable AI deployment. It really all starts with the design node. This is where teams can easily build, experiment, and refine data pipelines and models in a collaborative environment. From there, the deployer ensures a smooth and efficient transition from development to production, eliminating bottlenecks that you may see by legacy softwares. With the automation node, you can schedule and monitor workflows, ensuring reliable execution of data pipelines and batch scoring at scale. And then for your real time analytics, the API node enables the deployment of scalable high performance models, high performance model endpoints for instant decision making. Governance is really at the core of Dataiku. We really can't express this enough. Providing built in compliance, audibility, and security. Plus with the fully managed elastic compute, supporting Python, Spark, R, DataIQ delivers both flexibility and power for the enterprise AI. The connected ecosystem ensures that teams can move from data to AI driven decisions, faster with efficiency, scalability, and governance all built right in. Let's take a look at or let's take a deeper, but more of a high level look at the, governance side of things. Because here at Dataiku, we've noticed one of the biggest challenges in scaling AI is maintaining speed without sacrificing that oversight. Dataiku's governance capabilities come in to ensure projects are standardized, compliant, and production ready at scale. With the unified framework and workflows, DataAccu provides prebuilt structures that help teams standardize AI and analytics projects, reducing inconsistencies and improving efficiency. Next, the value and risk qualification framework allows organizations to assess the potential impact and risk of models before they go to production. This insurance project are both valuable and responsible. The model bundle and sign off process prevents unapproved or underperforming models from reaching production, reinforcing audibility and trust in decisions. With the centralized model and bundle registry, teams can easily track all models and their performance in one place, making oversight and iteration simple. Bringing it all together, Dataiku's centralized analytics and AI portfolio control towers, we like to call it, AKA the govern node, offers a high level view of initiatives giving IT leaders and executives the visibility they need to ensure projects are delivering business value while remaining compliant. Okay. Another big thing we like to, express here at Dataiku is the ability for collaboration and the reuse baked in. Right? Don't reinvent the wheel on things you don't need to. So one of the biggest inefficiencies of AI and analytics projects is obviously the constant reinvention of the wheel. Teams begin, building similar workflows from scratch instead of leveraging existing best practices. Data actually eliminates the friction of embedding collaboration and reuse into the core of the platform. Teams can store and share a line of business use cases, automation best practices, and integration examples. Instead of rebuilding solutions, teams can quickly access proven templates and promote them across the organization. Last but not least, we want to power AI every day through seamless access to Cloud services. At Dataiku, we make AI accessible to everyone by providing seamless access to Cloud services, ensuring teams can scale their analytics and AI projects with ease. The AI then to end data and AI platform integrates directly with Cloud infrastructure, allowing you to natively use Cloud services for data processing and machine learning. Collaboration is obviously key to successful AI and Dataiku enables teams to accelerate collaboration on data driven projects, bringing business and technical experts together in a simple intuitive environment. As AI adoption grows, governance is critical. Dataiku ensures you can govern AI projects and safely with built in compliance and security features to mitigate risks. Enough of slides. I'm pretty sure we will make sure you guys have the slides after this along with the recording. Let's hop into an actual demo. Let me close out of the slides real quick and let's hop right into Dataiku itself. What we are starting on is going to be the launchpad. This is going to be your central hub giving you a unified view of all your Dataiku nodes. Unlike Alteryx or other legacy softwares where deployments environments can be fragmented across different machines and servers, Dataiku is really putting everything together in a single easy to navigate interface. From here, IT managers can do everything from manage extensions, plugs in, change connections, manage code environments, to really even ensure performance across their entire ecosystem by checking usage and monitoring. Here right off the bat, we can see all of our design and preproduction nodes, our production notes, and then lastly, our governance note. We're going to touch on most of these today in today's demo. But I just wanted to give you a high level overview that it is all in one location and you can access each of them straight from this launch pad. The first thing we're going to do is we're going to enter into the design node and we're actually going to go to the local deployer. What I'd like you to think is as an IT leader, you're coming in in the morning, you're sipping your morning coffee, right? Getting ready, ready for your day and you're thinking, okay, I'm going to go in and I'm going to monitor how all of my projects in production have been doing. I can come into this monitoring and we can see that we have both our data Iq projects as well as our API endpoints and we can see we have a couple of errors within our data Iq projects. If I dig a little further deeper in, I can see most of my projects seem to be doing pretty well, but we do have a few errors in a couple of different other projects. The project we're going to be dealing with today is going to be this web demo CLV. As you can see here, the deployment status is healthy, the model status within the project is healthy, and the governance status is healthy. What seems to be causing an error is an execution status and that there seems to also be an error within the actual data itself within the project. As the IT leader, you can say, okay, let's see what seems to be going wrong with this execution status. Simply by selecting selecting the execution status, I can see the scenario, AKA the automation of the flow, which scenario is causing the error. Straight from here, I can click into this scenario, and it's gonna take me to the actual scenario run of this workflow on my automation node itself. We can see the steps that it's currently taking, and we noticed that it's actually failing due to a verifying of rules or run checks on a dataset. Straight from here, I can click into that error, and I can actually see the details of that error and why it is occurring. So we can see here I have a data quality rule on this dataset, and it's simply saying that I if I see any empty values in any of these four columns, let's make sure we're not running the rest of the workflow because whatever I'm doing with this dataset may fail. So I can see that I have a 36 empty values in my IP column. Straight from here, I can get then I've now diagnosed my issue. Right? And I can immediately start to work backwards on how to begin fixing it. So I'm gonna close out of my data quality. And right within here, I'm gonna actually open up the actual deployment of this project. Here, all within, once again, the same interface, we can actually see that this project has been failing for a little over a week now. So I can go in there, and the first thing I may wanna do is actually turn off that scenario. So I've now turned off this automation, and I say I have saved and updated because maybe we don't want the project to continue running without the the fix in place. So I've turned off that scenario, and now I've gone and I've said, okay. Now I need to die I need to figure out why this we are running into this data quality issue, which is then causing a, which is then causing an execution issue. Instead of working on this in the actual deployment node, I'm going to travel to the design project itself. Once again, all within the same interface, not downloading workflows onto local machines or anything. I'm simply selecting the design project and going into the project itself. Here we are now on the design node as you can see denoted in the top right hand corner, and we see the project homepage. The main thing I want your eyes to look at on this project homepage is going to be this timeline on the right. All changes are backed in Dataiku by a Git repository which allows for excellent tracking and version control of any project. So this might be the first place you may check, as an IT leader. You may say, like, okay. Let me find that dataset and let me see if someone's gone in there and made a change to that dataset. As you can see here, you can see me, myself. I went in I was when prepping for this, I was changing a lot of different things. So you see mainly me on the timeline. You would not only just see yourself, you would see anyone else that worked on this project during that time. So I can go to the flow, and that dataset that I was looking for was the new entries dataset. Right at the top, I can then search for it. It's going to take me straight to the dataset and I can now select into that dataset. Here you can see that the data quality issue is occurring on the design note itself and we can start to begin to work backwards. From here, I can scroll through my data and I can see that some of my IP columns are missing. Let's try to think logically about how to do this. Instead of working backwards, let's actually use the column lineage that is available to you within Dataiku to see if we can diagnose where this problem is beginning. Right on the right here, I can check the schema of my dataset and I can see the column lineage for my IP column. I can start to work backwards and see my full end to end flow and see where the data is coming from, how it's transformed, and at what point the issue occurred. As I begin to work backwards, I noticed to myself, okay, there doesn't seem to be any point in the workflow that this seems to be causing an issue. Maybe it's the actual dataset itself. From here, I can go to where this data is being brought into Dataiku, and I can immediately click on this dataset. When I open this dataset, we can see that this web data is missing France data. I went in here and I removed some IP address to basically make it so that let's think and let's pretend that all of our IPs from France, weren't able to properly be brought in into our source system. From here, let's pretend I know I didn't do that myself and this is being caused an error upstream. I can go in here and I can then discuss with people about what's going on. I can say IP missing. From here on the discussion, I could then call out someone on the team that may know how to fix this. So for example, I can go ahead and I can say, Chris, you know, he's my my, data engineer. He can go in and he can look at why, you know, our IPs aren't coming in in certain locations. So I can say, you know, check why we are missing IPs. Right? So I've now started this discussion. Chris Canal is now gonna be notified, and he can go in and he can take a look at this. So from here yeah. Let's say he's gone in and he's now diagnosed the fix and he's found the fix. Here's our new dataset that now has all of our IP, data. Let me go ahead and make that fixed real quick. From here, I can just replace my two datasets. I'm going to go into this tool and I'm going to change the input of this dataset and I'm going to go ahead and grab my web data that now has all of my IP items. So, I've now saved that. We can go back to the flow and we can see we've replaced that dataset. So the issue should now be fixed at hand. Right? If you use column lineage, we've, talked to the data engineer. We've said, hey. Please go ahead and fix this. We've now fixed it. Last but not least, what do we need to do? Right now, we need to go ahead and we need to bundle this back up and then redeploy it so that it is now back into production. To do that, very simply, once we have a completed workflow, we can come up to the top here and we can choose this bundles option. Through the bundles here, you can see we've actually already created three separate bundles. All of these were packaged at a specific date and time and then pushed to the Deployer. We've made the change, we've now fixed it. Our data seems to be looking a lot healthier. Now we can go and worry about creating an additional bundle. From here, I can say create a bundle. I'm just going to go ahead and I'm going to label this V4. Just the version after the latest one we have, and I'm going to go ahead and create this. When I create this bundle, it's now packaging up that entire project. Everything from the datasets, the data connections, the transformations, column lineage, everything. It's one for one, just moving from a design node to now you're either preproduction or production node depending on your organization setup. From this bundle, as you can see, we've now had a bundle, but it has not been published to the Deployer yet. Let's go ahead and let's publish that to the Deployer. I'm going to publish this to Deployer. It's going to give me a little heads up that it's actually occurring. I'm now going to publish it and I'm going to open up the deployer. We are back into that deployer section where we actually started this demo. We're not at the monitoring section yet. We'll get back to that. But this is the landing page for the deployer when you're actually pushing up a project. From here, right, the analyst or the data engineer, whoever is the data scientist is now uploading this into production, but then click deploy. They have two options. They can create a brand new deployment or they can update a deployment that is already being occurred. I'm going to go ahead and I'm going to choose update, and I'm just going to update the deployment that our v three bundle had previously been on. I'm gonna go ahead and hit okay. It notices that there was a previous deployment of this project, just a different bundle, and I'm gonna hit okay. Here's where Dataiku, really changes the game in which how, you know, we decide to govern things. Right? If you're used to Alteryx or some of, the other legacy softwares out there, you would notice that governance becomes something that has external documentation or manual sign offs. Right? Well, we build that into the platform here. So as you can see, without this bundle being sort of signed off and reviewed upon, I actually cannot, send this bundle into deployment production. Right from here, I can go ahead and check the governance status. This is going to bring me into the final section of this demo, and that is that the actual govern node. From here, I'm actually going to go back to the homepage just to give you guys kind of a high level overview of the govern node. From here, you can really keep track of all, governance initiatives, business initiatives, projects, and anything within the project itself to make sure that IT has oversight of what's going on, not only on in your production node, but also in the design node itself. From here, I can go ahead and we can start with business initiatives. As you have business initiatives within your organization, you can govern those. For example, demo projects. Then within that demo project is the project I've been showcasing to you guys currently. Within this project itself, I can go ahead and I can open up into that project, and we can see, a few things. Right? We can see, where in the lifespan the project is at, so exploration, qualification, and progress, whether it's invalidation. We can see all of the, related items to that project. Picture it like an umbrella and everything falls within that project. Most importantly, you'll see that v4 bundle that I've uploaded automatically has been added in here. It's in the review stage. As you can see, v one and v two had already been retired. So we're gonna worry about reviewing v four here in a second. You can see sign off for viewers and approvers, qualification snapshots. One thing I'd like to show you, before we actually get into the review is everything that you can actually keep track of within this cover notice. You can see obviously just basic metadata of the project. But one of my favorite thing is the actual related objects to the actual project itself. Within here, within the govern node itself, as you're trying to keep governance over a, not only once again, a production project, but a project actually in the design node. As an IT leader, you don't want to lose oversight during the design phase either. Here you can see everything from the bundles to the saved models, whether this is LLMs, machine learning models, agents, whatever it may be, knowledge banks, and then last but not least is your actual dataset. All of this can be controlled and governed within this governed node. Let's go backwards once again. Right? So we're in this project. I've uploaded that v four. I'm looking for it to get reviewed, signed off, and then approved upon so then I can actually put it in deployment. Right? So I can come to my v four bundle, and I can open this up, and I can click over here in the review side. From here, I'm now asking someone to actually sign this off for me. I'm going to request feedback and I can request feedback from a bunch of different groups. Maybe I need to get feedback from IT first, risk and compliance, the business leaders. I'm going to request feedback. Let's pretend I've gotten feedback from all those leaders. I'm good to go. And last but not least, I can request my final approval. So I'm actually going to delegate this. And I have Chris here, a member of my team. He's kind of, on the background right now. He's actually gonna go ahead and he's gonna approve this for me. So I've delegated that to to Chris. I'm gonna request final approval. I'm just going to choose Chris. He's going to get notified, but he's going to be the IT leader in this, or the data science leader, whoever it may be. He's going to get notified. He's going to go ahead and he's going to approve it. As you can see here, he's now approved this in the background. He says best safest project ever. Awesome. Good to go. Now that this has been approved, I can now go back into my Deployer. Let's figure out where that was. Let me refresh this real quick just to make sure. I can now deploy this. I'm going to update my bundle. As you can see, it's now taken us to the settings page where I can then go and return that scenario on that I turned off earlier. I can save and update, confirm. Now my project has been updated into my deployer node and hopefully tomorrow morning when this runs based off of its automation schedule, that issue should be fixed. Right? So, all in all, what did we see exactly? Right? We started with monitoring all of the projects in production. We saw an issue that was related to a data error. We then worked backwards, went into the design node, used column lineage to then figure out where that error first started. We'd found that in the source system itself. We'd made note to tell the data engineer to fix the issue and then re integrate the the proper data. We integrated that proper data, rebundled it, we packaged that back up, send it up to the deployer where it was then, signed off and okayed by our IT leader, Chris. Now we have it into production as a V4 instead of a V3. Kind of an end to end cycle of what it may look like to diagnose an issue, fix an issue, and then re upload your project bundle into deployment all within a single interface and end to end platform, which is Dataiku. Thank you. That is the demo portion. From here, I think we have some time for some questions, so feel free to, put your questions in the q and a section, and I'm happy to kinda answer anything that you guys may be curious about. Any more questions, guys? I know that we answered a lot, while you were doing your walk through Cody. So maybe we'll give couple of seconds for people to drop in anything for the time about. Going once, going twice? Yeah. The, demo and recording will be shared. Yes. Yes. Okay. So I don't think we have any more questions from the audience. So I guess that brings us, to the end of today's session. Thank you all for joining and for the great questions throughout. A huge thank you to you, Cody, for the demo and insights. Thank you. Again, as Cody mentioned, a quick reminder is that we'll be sending out a follow-up email with the recording and slides. Oh, a question just came up, by from Kaushik. Is it possible to deploy an open source room for a POC? I think that's a question to you, Cody. Oh, I'm sorry. What was the Oh, is it possible to deploy an open source version for, POC? Yeah. Yeah. So typically what we like to do, for our POCs is going to be, using Dataiku Cloud. So obviously, Dataiku can be installed on your, cloud infrastructure on prem or in our own Dataiku software as a service. So most of our POCs we tend to do, right within the Dataiku Cloud just because it is a lot less set up rather than, you know, deploying in your, in your cloud infrastructure. But yeah. So does that answer your question? I guess that does. Yeah. Kashik, if you have any follow-up to that, please let us know. But I think that's pretty clear to me. Yep. I think there's actually one more question. So I saw it in Seema tea. When you say shadow analytics, can one understand how prevalent it is? Is it? How much of a threat? How big is the iceberg under the water? Yeah. Absolutely. And I think organizations struggle with that, especially organizations that, you know, are, you know, deep within their life cycles with some of these older softwares such as Alteryx. You don't realize how much insight you're missing on things, until you really deep dive in. It depends on your organization. It depends on how integrated that software is within your organization. But, yeah, I mean, I've seen the iceberg be pretty large. Right? And and once you kind of start to discover what's, you know, underneath that bottom part of the iceberg, you really kind of really start to see all the things that you should have uncovered previously that you're not uncovering already. Is it possible to read PDF file and convert? Yeah, absolutely. So there's actually multiple ways you can do that within, Dataiku. We have, OCR plugins, which just read directly from, PDFs, or you can actually implement, you know, some of these newer use cases, revolving LLMs and AI to actually read information from those PDFs themselves. I think the next one's from Adam. Do you equate shadow analytics to dark data? Adam, can you define dark data for me on that one? I think a lot of people refer to refer to it in different ways, but I think I think what he means is more like lack of visibility, I think, is what he mean. Like, a black box is how I also refer to it before this woman are, but I yeah. I think, Adam, if you wanna, elaborate on what you mean with dark data. And Seema's question is, what are the preventative measures from a design point of view? Yeah. So the equating shadow analytics to dark data, I mean, is that part of the issue? Yeah. Absolutely. Right? I think having insight into not only your datasets, but what's happening with those datasets is clearly important. Taking it back to kind of what it showed in the government note a second ago. Right? It's even it's nice and it's, you know, really useful when you can govern projects, but the fact that you can actually govern the actual datasets themselves. Right? A project is gonna be nothing without that data. You can you know? So from governing and and having insight into your starting point is just as important as having, governance and insight into the rest of your actual, platform and the transformations and everything that are occurring. Yeah. And then what are the preventative measures from a design, standpoint? Yeah. I mean, you know, from the design standpoint and being preventative in it, are are we talking as far as, like, how to design projects or from, like, an overarching umbrella of just governance in general, Seema? I'm curious for myself. What about, like, from an overarching governance in general? Yeah. Yeah. So, really, that's gonna come from, you know, your your IT team or just business initiatives together, in general. I think it's something that has taken a back seat in analytics, you know, just from how quickly things have moved from not only just, you know, on prem source systems, you know, everyone, you know, storing systems in boxes underneath their computers to all this cloud storage now. I think governance at one point or another has kind of taken the backstage, during that transition time. But now as, AI is becoming a lot more prevalent, and, you know, people are starting to trust machines to actually do things that they want to be doing with their data. That governance is now kind of creeped up to the top of, IT leaders and just kind of organizations in general and understanding how to control the chaos of what's kind of going on, when it comes to their data. Yeah. Makes a little sense. Yeah. What Dataiku is really trying to accomplish, everyone, is we don't want you guys to be dealing with any black black box at all. Right? Whether it's your data, whether it's the transformations, whether it's the advanced analytics you're gonna be doing within your organization. Dataiku is meant to be kind of this orchestration platform that kind of sits on top of everything, and it's really kind of showing you what's going on, when it's happening, how it's happening, all within a single platform. Yeah. I think see what's the other question you've already answered. Right? When we see shadow analytics, I wanna understand how prevalent is it. Oh, how to prevent those shadow analytics? Yeah. I mean, I think, honestly, the first step, right, is understanding where shadow analytics may possibly be happening in your organization right now and taking a, high level overview of why it's happening. Right? I think, like Adam just said, in the chat, I I totally agree. Right? It's sometimes the last aspect to be considered and only ash after ish issues have been identified. So identify the issues. Right? Identify the issues, work backwards, reverse engineering, figure out, okay, what can we implement in our organization, to make sure that, you know, governance is now taking a four c and really kind of driving change and adoption within your organization. Yep. Very well said. I think that brings us to our last question. I mean, that was that was our last question. Anyone else have anything while we still have Cody? Looking at the chat, the messages. I don't believe that there are any more. So I guess, again, that brings us to the end of today's session. Again, thank you everyone for joining. Thank you, Cody, for the demo and the insights and for answering everyone's questions. Along with the recording and slides, we'll also include additional resources to our follow-up email. This will be great, especially if you're looking to dive deeper into how Dataiku compares to desktop based tools like Alteryx or what a modern governance first analytics stack can look like. And if you're thinking ahead, don't miss out on our upcoming webinar on AI agents by our very own Florian, who is CEO and cofounder of Dataiku. It's happening this Thursday. It's one of the hottest topics in enterprise AI this year. And so this is definitely a session that you won't wanna miss. Other than that, thank you again, everyone, for attending today's session, and we'll see you next time. Have a great day. Thanks, everyone. Thank you.