00:00:00

<silence> Awesome. Good afternoon everyone. Thank you for joining. I think, I think we are good. Let's get started. So my topic today is gonna be architecting a Optimal finops platform. Um, how many of you are familiar with finops or work with finops today? I see some show of hands. How many of you have never heard of ops? Good. So I'm not going to spend a lot of time talking about what ops is. I'm just gonna assume that, uh, we know what OPS is. I'll probably spend a few minutes just talking about what the, you know, concept behind Synopsis. But other than that, we are gonna jump right into the platform side of things and talk about how finops and DevOps sort of come together. My name is Ajay Shma. I head up the platform engineering team for ThoughtWorks North America. Uh, we have our ThoughtWorks colleagues from Netherlands here.

00:00:50

Um, so if you need any, um, question, if you have any other questions, need any answers to any of things that we are gonna talk about, talk to any of us. Anyway, let's jump right in. Um, so one of the things that we should always be thinking about when you talk about Synopsis, um, you know, we all have cloud cost optimization problems. Anybody who's using cloud has a cloud cost optimization problem. It is not something like the previous speakers we're talking about, not something that's gonna completely go away at any point in time. It's always going to be there. It's a question of how do we actually approach that. So that's basically what I'm gonna do today. If you again, need, um, some crash courses in finops, the best way to get that is check out finops.org. Uh, that's actually, um, in a Enos foundation under Linux Foundation, uh, you can get a lot of information on that.

00:01:39

So what are we gonna talk about today? So, I wanna first talk about like every other conversation, like Gene talks about, talk about what triggered this whole idea of creating this Enos platform. And then we'll talk about what the overall problem definition, and they will talk about how are we solving the problem, and then we jump right into the platform capabilities. You know, why is this, why should this be a platform versus bunch of automation that we do? Uh, then we will actually try and tie this back into the whole DevOps problem, right? Because we have the, our pipelines and our engineering effectiveness workflow is well known these days. So we need to really try and see how can we actually marry those two things and see how can we make sure that your developers are enabled to solve this problem? Not as an afterthought, but as you build your, you know, your products.

00:02:23

And then we'll talk about some takeaways. So first thing, how did this start about, right? So if you really look at what is happening within the industry, uh, you see that significant amount of cloud cost optimization problems are being solved in a tactical way. There is no strategic way of solving it. Again, this is something that we understand, we see on a day-to-day basis, having worked with about like, you know, a hundred clients who have been doing this on the, on the cloud for the past seven to 10 years, but also from the industry, from, from what you really read on the literature, right? But you also find that fact that, um, this is not a very unique niche problem. 'cause if you take the digital natives out there, about 85% of them are on the cloud, and they have this problem, and they have understood there is a problem with the cloud cost optimization though.

00:03:10

So everybody's trying to shift left, just like every other best practices within DevOps. Everybody's trying to shift left, meaning you're trying to solve the problem sooner than later. That's great, right? But you also know the fact that, uh, this is a fairly conservative estimate. If you really look at it, about 30% of the cloud costs are currently, you know, wasted. If you think that you are, you have done better than that, definitely take a look at it. You might have, but there is a really good possibility that your wastage might be more than 30%. But, um, the, the other fact is that this is the most interesting thing for me, right? If you look at a particular year through which you actually budgeted everything for your cloud, and you're, you ended up making some optimizations, then go into the next year, you continue to under underestimate and your, you have, your budgets still not aligned with what you really want it to have.

00:04:01

And, um, so how do, how do people solve it? The biggest ways in which people try and solve it is you go out and buy a tool or a platform, that's great, but tools and platforms are in solving the problem for you today. Because any tool that you buy out there is going to perhaps solve little bits of your problem, but it's mostly going to tell you what the problems are solving. The problems are still with you, with your organization to do it. So that's where you find that if you buy a lot of tools, if you buy a platform and try and put together a finops team that isn't going to actually reduce the cost, it might actually end up increasing the cost. So let's talk about, um, you know, what, what this means. So this means that organizations typically need a holistic solution, not like an isolated, localized solutions, right?

00:04:49

I mean, that is something that we see all the time. And we also see that, uh, this is, this holistic solution is not just a bunch of tools and process. This is about building the right kind of solution contextually as to what that works for you. So now what's new about solving this Enos problem as a whole, right? This should not be that different because it's just the same, right? So you understand the problem, you try and see that, you know, you extrapolate the data, make sure that you get the right kind of data to say, okay, where my problems are when it comes to, comes to cloud cost optimization, then you try and put together some kind of phase work plan, go ahead and optimize it. So that's great, right? I mean, this is no different from any other problem solving that you have within the finops world. I'm sure you have seen this. There is this whole cycle called infor optimize operate phase. So essentially you try and collect the data, you make some fixes, and then make sure that that's your BAU. And that's great too. Um, because this is, and again, if you haven't seen that, um, uh, flywheel, don't worry. I mean, because that's very, very easy and simple idea. And I was just simply talking about infor optimize and operate based on how you might actually solve any problem. And you find

00:05:56

That there are lots of platforms out there, right? I mean, you might have, if you are in this field, if you are using cloud, I'm sure you have had some tool, you have used some tools to do this optimization. So essentially the tools that are out there, there are plenty of tools out there. Tools that are out there essentially gives you the reports and tells you the, you know, recommendations on what you should do. The next thing to think about is that, okay, that's, that's all great, but how can you automatically make some optimizations? There are some other tools out there that actually looks, looks at some of the backend catalogs of your cloud providers, gives you some a base rate optimization idea. So of course that saves some money. Uh, let's say if you are trying to save, you know, about a hundred percent of your wastage, uh, that probably gets you up to 30% of your wastage, um, in a savings.

00:06:44

So now comes the biggest problem. So when it comes to really trying to create a solution that will actually solve your problem, the overall wastage problem, you still don't have a solution that you can go and buy, go out and buy. So if somebody tries to sell you a platform or a tool saying that here is your solution for your f now problem, buy this and your problems are solved. I'm, I'm sorry, that doesn't solve the problem. You know, if that solves a problem, we should talk. I don't think it does. So, so the way I, I typically talk about this is that there are three Rs of this ops, right? So this whole idea of reporting recommendation and remediating. So this is fairly simple concept for us to keep in mind reporting and recommendation ECF said, and there are lots of fantastic tools out there that actually tells you what the problems are and what you should do remediation to solve the problem that requires more contextual information.

00:07:37

And that, that's what we are really gonna talk about. So, uh, so as we talk about reporting and recommending, one thing that you should be thinking about is that, uh, I keep referring to this as accelerator platforms. So these accelerator platforms, there are several platforms out there. Just to give you a flavor for a couple of these things. You might have heard of a tool called cube cost. I'm, I'm sure you would've heard of things like Cloud Zero and cloud gener, and, you know, uh, cloud checker, there's so many tools out there. What these tools does is, I mean, what they promise and what they do is, is great. It tells you what your problems are and tells you how to solve them. So for example, if you take something like cube cost, um, you see a screenshot here, it tells you a lot of information about, okay, here are the kind of, uh, uh, current usage when it comes to your Kubernetes clusters.

00:08:22

Here are the, the current usage of your, your overall, uh, cloud profile. And it also goes to the next step and tells you, here are the things that you should be doing. So it tells you things like, okay, here are the unassigned resources, here are the amount of, you know, uh, areas where you have some abandoned workloads. So these are the kind of things that you should be doing. So you see that the difference between doing it and telling you what to do. So this is great. This is the primary information that you need. But now the next step beyond that is once you have it, you translate that to realizing that remediating this problem. So another set of accelerator, um, tools out there. So, um, this is from a, uh, from a tool called Prosper Ops. I'm sure some of, at least some of you have heard of it, there are lots of other tools like, you know, SPOT io and IBM Turbonomic and a lot of other tools that tells you, uh, take takes it to that next level.

00:09:15

That next level here is how do you actually look at the current rates that are available for a, for your cloud provider? How do you automatically apply that without you having to jump through any other additional cos? So this is great too. This gets you that, you know, what I talked about earlier, right? Going from zero to 100, you gets you from zero to 30, you still have that 70, uh, no, 30 to 100 that is remaining out there for you to continue to optimize. But these are really good tools if you want to just get started and see some value, but that's where it actually stops. So the, uh, another one from a sustainability point of view, this is something that, uh, we built, um, you know, in, in collaboration with in a few, few other foundations. Uh, this is called cloud carbon footprint CCF tool.

00:10:02

And if you want to check more, more about it, let check out, uh, cloud carbon, uh, footprint.org. Um, essentially what this does is this is essentially a set of APIs that works with your cloud service provider, gives you information about your sustainability user, your carbon footprint, and how can you actually optimize that. So very similar to the other concept we talked about in the accelerators, but here it specifically focused on the sustainability side of things, and it gets a lot more practical when it comes to that. So it tells you like, you know, if you use these set of resources for these kinds of activities and these kinds of nodes, here's the amount of carbon footprint that you're leaving on the table. Here's the amount of sustainability that you are proud of, sustainability goals you're not meeting. So that's the level of optimizations that you could do.

00:10:47

Again, this does not do it for you, tells you so that this is all great, but, uh, one question that we always ask is like, whose problem is it to solve? Because when we talk about finops, one thing that we keep hearing is that, um, you know, uh, it is a finops team's problem or it's a technology problem or a product problem, or, uh, a finance problem. So let's say if it's a finance problem, so it essentially, you know, you think of it like you have, finance has all kinds of requirements with respect to doing a better budgeting. You know, you forecasting, you know, so you have all kinds of cost controls come through that. So it's a finance problem, you know, it is absolutely right, but could it be a product problem too? You know, not everybody thinks of it as a product problem, but you know, it could be a product problem, right?

00:11:31

Because product is thinking about it from the point of you have your cost of revenues, your cost of goods sold, trying to make sure that, you know, if you are actually selling a product, you wanna make sure that you are the kind of, um, money that you're spending to actually sell that product. You build that product and sell it isn't, uh, more than what you're really trying to make out of it. So that's great, but eventually it turns out to be a technology problem, right? Because, uh, you know, the onus is on the actual developers, the who are actually building the product to make sure that you understand what your challenges are and try and solve it. But the answer to the real question, whose problem is it to solve it is that it's everybody's problem. And this is what makes it so complex, right?

00:12:14

So, um, again, I said I'm not going to go too much details into what ops is, but I just want to give you a very high level view of this so that you can understand some of the ideas I'm talking about with respect to platforms, right? So there are six key concepts when it comes to finops. Uh, there are very simple concepts. Essentially, the idea is that these principles drive the whole idea of cloud cost optimization. The first one, of course is the idea of the ownership of your usage, and then you need to make sure that you're collaborating. So we just talked, uh, talked about there are multiple access in which the ownership lies. So you wanna make sure that the clear understanding of the ownership is there. And the idea of centralized team, you have heard of think concepts like finops team, right?

00:12:52

That actually drives it. That's a glue team that really brings all these people together. Then, um, a common misconception when it comes to cloud is that people always talk about, okay, I want to do cloud cost optimization, so I'm gonna save money. So let me go to the cloud to save money. You know, that's an a really unfortunate way of looking at it because finops is not about saving money, it's about making money. It's about increasing your business value. Um, so, but again, it's easier said than done, right? People are going to say that that's all great, but I still wanna save money, right? So the other big aspect of it is, you know, your report accessibility and timeliness. How can you make sure that when you are generating all these reports, how do you make sure that it talk talks to each other and it gets the right kind of information there and eventually the whole variable cost model of the cloud, you know, which is something that is super important.

00:13:41

That's one of the main reasons why people go to cloud in the first place. So, uh, now let's talk about like, why is this remediation problem something that has to be solved through using a platform approach? Why are we trying to do this with multiple clients of ours? Um, the primary reason for that is that if you look at, uh, any data that's out there, this has come up through a lot of the work that we have done with so many of our clients across all the different industries that we've been talking about here today. But beyond that, if you look at data, do finops.org, you can find this, um, data collect this very similar kind of data collected from 10,000 practitioners of finops. So what you really find is that there are multiple things to consider here. Number one, the data that is actually used to make decisions.

00:14:25

Number two, the automation priorities. Once you have the data, you're trying to do some automation, what's the priority for your doing your automation? The third one, remediation. If you're trying to do the the remediation problem, what's are the priorities for that? Then eventually you solve this problem. So what's the level of maturity that you have? There is an incurrent misalignment between each of these activities happening today. So this is based on the data that you can really see, or@datadoobs.org, right? So when I talk about data, what are the kind of things that I'm talking about? So fairly, uh, obvious things, your cloud, cloud utilization data, your finance data, and you know, all the kind of data that you get from various data sources, that's really gonna drive the decisions. Then you think do things like automation. So if you are looking at things like anomaly detection, you, you see that if your cloud cost usage is lesser than what you expect, it's not always a good thing, right?

00:15:17

Because that might mean that your customers must be struggling, you know? So if it is more than what you're expecting, it's not always a bad thing either. So you really need to know what your baselines are and try and make sure that your anomalies are understood properly. So similar to that, things like usage reports, and you know, utilizations, you see that the priorities of what your automation is happening or how your automation is happening is somewhat different from the data, a availability that you have. And this is also interestingly different from your remediation priorities. So from a remediation point of view, we know that anybody who has played around with finops know that the number one thing is tagging, right? It's all about tagging, making sure that you know what your resources are used for. Are you tagging your resources correctly? And if you look at it that way, that is what the priorities of remediation are, which is somewhat different from your automation priorities, which also you saw that from some is somewhat different from your data priorities.

00:16:13

Now, let's look at the actual maturity of these things. So you do all these things, people tell you, here are the things you should be doing. You build your automation, you say you're going to do all of these things and you did it. Now, the maturity of what you're really trying to do is somewhat different from that too. So here, the number one thing that from a maturity point, if you see, is the unit cost economics. So most people who do use the cloud know that there's unique cost. Economics is one of the driving factors. Then we have develop some level of maturity there. For those of, you're not familiar with the uni cost economics, fairly simple concept, uh, idea of trying to make sure that, you know, if your pro, if your product or a service or an application is using certain amount of resources, what are those, those resources, what are those?

00:16:57

You know, the, the money that you're spending for that. So you can see that there is inherent misalignment between all of these ones. Now let's jump into the platform capabilities as a whole. So this is something that should resonate with you because this has got nothing to do with finops. This is about the overall aspects of how you should really be building the platform. What are the capabilities that you're trying to do? So cognitive load, we talk about this a lot, right? We want to try and make sure that there is reduced cognitive load as you're really trying to build this FENOs platform. Um, you know, those of you have a team, apologies, you know, quite familiar with this idea. So efficiency, um, making sure that you are, you have shared responsibility when it comes to the shared responsibility model. When it comes to this.

00:17:39

Not one team is trying to make all the decisions and make all these things happen. This is something that has to happen across the board. The third key category agility, making sure that, that you, there's the proper level of discoverability here, you know, because this is becomes huge when it comes to the complex services that are being offered by every cloud service provider. Uh, the replaceability, this is another key aspect of this. So remember we talked about all these accelerator platforms that gives you reporting and recommendations, but what if you are really tied to that platform forever? If the, what if the licensing model changes? You want to have a model in which you can change those things, right? So those kinds of things, replaceability becomes extremely important. Compostability. So you can see in a, in a, in a bit the overall platform as a whole.

00:18:24

What we are really talking about is not to reinvent the wheel, it is to build on the kind of things that, you know, you have been using as an organization. So you wanna make sure that the platform, key platform capability has those compostability built in so that at any given point in time, you can actually start shaping the changing the shape of the platform itself, the most efficient way possible. So, um, now just, just to sort of look at looking at the building blocks of what this platform is, right? So the first set of building blocks would be data related things. So where here, what you're really talking about is that, uh, there are so many different reports coming in, how do you actually make sure that there is a timeliness and correlation of the data, right? Then the second thing is your organizational idea.

00:19:08

So essentially trying to make sure that, do you really have a DevOps culture? Do you really enable your developers to do the kind of things that you want them to do? Or are you expecting somebody else to make those decisions while your developers are building your product, right? So that becomes a huge part of how you actually, you know, build this third one pipelines. Pipelines are essentially a commodity at this point. So when you're really trying to build the pipelines, how can you actually incorporate some of these things into your pipelines so that this doesn't turn out to be an act, an afterthought that you end up doing after you start finding that, oh, here's a bunch of problems I need to solve, so let me go and fix my pipelines. It should not be that way. It's something that you have to really start doing from day one.

00:19:48

So your overall operating model of platforms, you know, so this is where you really start thinking about like, um, you know, could you actually make your developers a lot more self-sufficient? You know, could they actually self-manage the resources? You know, could you actually have a more product driven way of doing things? Because all platforms, as we all know, right by now, has to have a platform product model, right? This is all platforms are built as products, especially if in our, uh, platform would absolutely be built as a product. So you need to have that product thinking coming in. The last thing that we just spoke about, and I'll talk a little bit more about is the sustainability side of things. Again, doing the right thing should not be an afterthought. Doing the right thing with respect to sustainability should not be something that you'll, you'll do when you have time and money to do it.

00:20:34

You know, you can actually have it both ways. So that's something that we really have to think about as you build this platform. Now. So talking about the concept of sustainability of the Green ops, uh, we don't really look at this as a completely different entity than finops, right? For us, green ops is a part of finops and it has to be one, and it is, and and the reason for that is, what I've done here is I tried to map the whole green ops cycle into the finops cycle. We just spoke about your infor optimize and operate cycle. So here what you do is exactly the same thing. So you get your sustainability data, understand your targets, and then you try and make sure that you have some, some kind of, you know, sustainability aspects built into your IAC and and pipelines and integrate with all your third party tools, whether it is a CCF tool or any other tool that you use.

00:21:24

Then you try and make sure that you have your alignment with your ESG and then eventually look at this with the right kind of automated governance. So you can see that the activities that go into implementing a proper green ops cycle is no different from how we would typically do in afin ops cycle. So it doesn't take an extra effort other than the fact that you have, you have to get the right data. Once you have the data, you act on it the same way. Now, coming to the actual platform itself, right? What does it look like? So the base layer of it, we know everybody uses some kind of cloud provider. Of course you have have that. Now, based on that, uh, you can see that there are some key components there that are a prerequisite as you build this platform. So those components that you see here on the, on, on the, uh, right side, right extreme side, you can see those accelerator platforms.

00:22:14

So these are those platforms that we talked about that provide you that reporting and recommendations. Then you have some, some kind of sustainability, uh, accelerator platforms there. You also have your observability platform. So this makes the assumption that you already have some kind of an observability platform in play. If you have it, that's great. If you don't, that is absolute prerequisite for you to build it. So assuming you have it, I'm sure all organizations have it, even if you don't know it as an observability platform, you probably have it. So that becomes another key aspect of it. So now comes the five key components of this, um, this platform capabilities. So those capabilities would be your, your core capabilities. That would be things like your resource tagging and entity right? Sizing, all the things that you know, typically from the point of view, what are the kind of things that you should be building to ensure that you get what you're, uh, trying to get in a automated remediated faction.

00:23:07

Second metrics, right? We talked a lot about it already, how to actually get that data and try and put, put it back into the, into the system so that it can actually start making those decisions by itself. Alerting and notifications, again, not an afterthought. This is something that some tools already does it, but how do you integrate all of those things? How do you bring it all into one umbrella? The policies around some of these things. So essentially, this is where cloud becomes really interesting and as a technologist, probably, you know, you may not think about some of these policies, right? These are, some of these policies sort are financial policies, your business policies, how do you incorporate that and make that as part of your overall platform? Then your automated governance. So if you expect somebody to, a team of people, whether it's a <INAUDIBLE> people or your finance people to sit around and really make the decisions on, okay, here are the kind of things that you should be doing that isn't going to happen.

00:23:56

That is not a scalable model. So you need to build your automated governance here. So let's not forget the fact that obviously when you're building this, you have to build this with an API first mindset, right? I mean, this is not something that, you know, you can manually manage at any, any point in time. You also really see that, I'm sure if you really look around in the industry, one of the things that you keep seeing is the fact that there are a lot of machine learning happening within this space, right? So there's a lot of AI based decision making that is happening, and we remember the, the tool that we talked about earlier that does the automated rate optimization part, that those tools already use a lot of AI capabilities. So having your platform, being able to understand what is happening with within the actual, uh, based on the data, what's actually happening within your cloud space, within your current contextual environment, it should be able to understand that and make decisions based on that.

00:24:48

Then let's not forget the cultural aspect of it, the people aspect of it, right? Um, and you might ask like, how am I actually going to do that? Um, cultural aggregators as a tool, yes, there are several ways in which we can actually do that. You know, you want to make sure that, um, anything that can be auto automated, anything that can be driven through a developer experience, aspects of things should be incorporated in here. You know, the, the talks earlier today, some of them covered some of the, the cultural aspects of it. Those are incredibly important when it comes to building this platform and making it successful. So, um, now let's try and tie this back into the overall DevOps lifecycle. So we keep talking, but I'm sure you've heard of this term of engineering effectiveness or been hearing a lot of late, the whole idea of engineering effectiveness is making sure that how are you making sure that your developers are more effective, so, and they're more productive so that your customers eventually get that benefit of doing this.

00:25:43

So the way I would look at it is a synopsis, essentially something that has to be built into every step of the way from day one. So it's not an afterthought. So how does that work? You're planning, you're defining your, you know, designing, defining, and making sure that you are, you're delivering your products each step of the way. You're going to have things that you can do and you should be doing when it comes to being able to, uh, do your, uh, doing things like, you know, making sure that you find the right kind of, uh, you know, uh, ca right kind of ba backend catalog, things to choose the right kind of resources, making sure that you integrate with your observability platform, making sure that it, you have your IAC built in such a way that you have your tagging and all the right sizing and all the things that we talk about becomes an automated way of doing things.

00:26:28

So eventually making this a continuous cycle, make sure that you don't really go and fix your FENOs problem after it happens, as opposed to that you're trying to make sure that your FENOs problem does not happen, right? So that's the way you should really start thinking about, and that's the only sustainable way of solving this problem for the long run. So, um, we can't talk about any of these things today without talking about a shared responsibility model. So, uh, again, this might, there's, there's probably a little bit of busy slide, but I just want to talk through that, at least a color coding there. So we talked about multiple owners to the space, right? So you have your product management, you have your, uh, finance, you have your development teams, you know, you have probably your third party providers. So having a clear sense of ownership is extremely key to make sure that this becomes successful, right?

00:27:16

Because when you have too many components in there, uh, and what we have seen is that this is the number one reason why any of EOPS offerings don't become as successful as you want it to be. So you want to try and make sure that this trade responsibility model is something that you'd start with at least. And, you know, if you have a better model in which you have actually tried to, you do do any of these things, uh, let's chat. I would love, love you to get into very specific details of each of those components and see what it does. But essentially what this is telling you is that, you know, you really need to make sure that there is clear understanding of the responsibility across the board as to for everybody who's involved in this. Um, so what are the takeaways? So we talked about why a platform for remediation, right?

00:28:00

So because we have lots of things out there for reporting and recommendations, hardly anything for remediation, doing it this way is going to solve the problem for you as opposed to knowing what the problems are. Um, we talked about some of the key capabilities of the platform. So this is, um, key for you to really understand that what are the kind of, you know, context that you have, what are kind of problems that you have, because it's not about really buying an off the shelf solution and using it, right? It's about understanding your problem and how do you actually map it to some of the key capabilities that we have. And, uh, then we talked about some of that shared responsibility model and make sure that how can you make sure that the right people who should be thinking about it are thinking about it and solving this problem.

00:28:43

So, um, so where do we go from here? So sort sort of to conclude, I wanna make sure that I talk about things like what's out there, what is still needs to be done, right? So this can be split into multiple areas, starting the strategy, right? So I think we talked about some of these things, like what are the ownerships, uh, ownership aspects that you really have to think about or what are the right kind of skill sets that you should have if you're building the partnerships with all these vendors who could possibly be selling these products for you? You know, especially on the reporting and recommendation side of things. What are the kind of strategic aspects that should be thinking about? Um, also on the, on the reporting and recommendation tooling there, what are your real goals? Are you just buying a tool because somebody promised that, Hey, here's a tool that's gonna solve the problem for you?

00:29:26

Because I, I gave, I show an extraordinary example of cube cost and talked about so many other tools out there. These are all great tools, but is this the right tool for you? So I think how do you actually make that decision? That is something that organizations still struggle with. Sometimes we get involved with them after the fact, after they have actually bought the license, and then you find that that's not the right solution for them. So that's probably too late for, for you to actually make that decision. So you need to be thinking about it. Uh, shifting left, you know, we talked, talked about this a little bit, you know, I think more from the point of your, of your engineering effectiveness. How are you actually going to do some of those things? Um, and the cost side of things. I know you really have to start thinking about like, is there a unified way of looking at your cost and usage?

00:30:09

You know, Enos Foundation have recently started working on some of these things. This is still an evolving area, so you need really start thinking about how can you actually do this, um, consumption. So we talked about tagging, continuing to have more innovative ways of doing. It'll be great. Uh, you know, as, as of I read, read some somewhere recently that, you know, the current maturity when it comes to tagging is about still 95% of the organizations don't have proper tagging, right? And eventually looking at the point of view of way stage, how can you actually get more informed targets and try and fix these problems? So, um, to conclude, um, you know, check out our website on, um, you know, some of the key aspects of this platform and how we are actually building some of these things. Uh, you might also be familiar with, uh, my colleague and chief scientist, Martin Fowlers website, check his, uh, blog out and know you might have some really interesting information on the platforms there. So questions, reach out to me. Uh, I'm here for the next couple of days. Reach out to my colleagues too. Thank you.