 We will get started with our next panel. I'm going to try really hard not to run as far ahead of schedule as we did this morning But we're just a couple of minutes ahead of schedule So I'd like to welcome our first panel of the day and I will introduce Jorge and I'll let him introduce everybody else So take it away Is it working? Yeah. No, it is perfect. All right. Thank you so much for being here We are thrilled because we're gonna be talking about a Topic that is I guess the bread and butter of platform engineering because we've heard some very inspiring talks on how to build your platforms or for approaches or specific technologies But we're gonna what we're gonna discuss today is really the the difficult choice of what do you actually do? Because you're not gonna build out your entire platform You're most likely going to have pieces that you buy from mentors You are gonna build other pieces and then you're also of course gonna adopt opens or so as we wouldn't be here So I am Jorge La Fiesta. I am the author of the Linux Foundation introduction to backstage course I've been while I was working with backstage for over three years so that means I've been very close to the platform side of things and Now I'm working to Rootley, which is an incident manager So it's slightly more tangential, but I'm still very interested in the area and today with me We have a fantastic lineup representing different industries from finance to startups So it's gonna be a treat today for the discussion. So let's start with some introductions Hello, everyone, my name is Tinhu Bai Hot. I'm from Saxo Bank. Saxo Bank is an investment Investment Bank offers online trading platform I work as a chief cloud architect and I'm leading a few platform Engineering teams, so I look very much forward to the discussion today with everyone here. Thanks Hello, everyone, my name is Edgars Petro-Rajas. I'm working at the Lego Group this toy company. Probably you all know I'm an engineer. I work with the Kubernetes platforms before now I'm working with API management and focusing on events framing platforms and Yeah, this subject is really really hot topic right now Had worked with different vendors and open source platforms. So really hope to share my experience with you Hello, I'm Lina Munaram and I'm a platform engineer at Chainalysis and Chanalysis are a data platform for the blockchain and Really what I do on a daily basis is try and make our engineers lives as easy and hopefully as Stream as possible Hello, my name is Victor Araujo. I work as an engineer in the developer experience team at Walt. Walt is a Leading and delivery service for food and more. We're present in 27 countries and we're part of DoorDash International I have worked for over six years with Kubernetes and different the platform and infrastructure teams and Exciting to be here discussing this. Thank you All right. Thank you for being here today And we're gonna have a little bit of a baseline of a question because as you know platform Even though we have a fantastic paper of white paper from the working group on what platforms are and capabilities It is still very much varies what a platform means for me and in my context in my company and My industry so we're gonna start with a little roundup on What does platform engineering mean to you or what actually is your platform in the one that you're working? So we have more context for the follow-up discussion that we're gonna have so who wants to start? Ding Hong I can quickly start Platform for us is a place where you can get your service. So it's like a catalog where you can get I Don't know for example a database or access to the database and everything around it We need to make it painless There's no need that for Amazon Azure or Google offering such great API and still need to go through ticketing system Fill out all the specification in order to get your platform Component so platform to us is a catalog where everyone can order things easily And it might not be as customizable like you have on the combination But you will have all the best practices and security measures compliance requirement building Sure, maybe I'll just I couldn't agree more with you on this one I think the key word for us is of course the cognitive load What we're trying to do is reduce the cognitive load. We're not our primary business is not Working with let's say it but still we have many many developers who would like to focus on innovation and you know working on things that they really care about and With the years we see what's happening that there's more more different cloud providers platforms to learn I think I'm making jokes now like how many CLI tools any of us need to know by this point, right? So I think it's really important that we make sure that developers are happy and that they work on things That they really care about Absolutely again, I agree and I think if you go back to to the word it's platform And I think the whole idea is that it should be there to support engineers and to be that platform that our engineers can use to build Build their software Kind of easily really and it is you know to reiterate It is about trying to make the right thing for the organisation the easiest thing to do and how we can help Do that as painlessly as possible As an engineer I used to think of platform in terms of the Technology and the teams that I was thrown in often in many organizations. There is a single platform team but what really Broadened my horizons is to learn about the product and business side of platform models where you connect Service producers with service consumers and then that transform transform the way I think about platforms and How I understand it now is that we are facilitators to connect these Service providers it could be internally and when you have this mindset It also allows you to scale in your organization from having a single team to having a team that is enabling teams that will come and take ownership over specific Areas that you want to provide to fulfill your Organizational mission with technology so that's what I would really like to Encourage developers to also learn about this product and business side of the business platform model Yeah, absolutely. I think everybody mentioned like this key Corner stores of platforms and why we are even talking about this concept. We have the distribution of ownership We have ease of use we have enablement for developers And then a very important part that might be driving this entire thing is the business value that we can Push through while enabling everyone to work And of course, this is a very difficult endeavor as you have experience in your own flesh And that also requires a lot of coordination and collaboration So how do you go about? Building a platform in practical terms. How do you identify what you have to build or what you have to? Just shop around for something that's already built or when you can adopt something on the open source and make it your own Like what's your initial step when you are like confronted with this blank page? There is no platform And we want to get somewhere. How do you start? Yeah Yes, as you mentioned, there are so many products out there. We can we can build ourself we can adopt or we can buy off in the shelf I think many companies probably face similar thing like just compare how much effort is to engineering these components Compared to how much does it cost to buy it? But for us we take both into account, but we're also thinking about an operational cost So when you buy something doesn't mean that it just works forever No, you still need to maintain it. You need to upgrade it. You need to customize it. There's so much to it so We focus a lot of our customization and scalability when this makes sense We we can buy things but if it's open source product, which has a great community support good documentation We also don't hesitate to just to adopt it I think the first step is to go to the drawing board and think about What you want to achieve with your platform in your organization because the business will be investing a lot of money and your engineers will be putting their sweat blood and tears to Make this happen. So Some common goals might be Velocity developer satisfaction it could be that This velocity will give you an edge in your business It could be security reliability many things, but I think the first step is to Think what you need to achieve and why this investment is going to be justified in the end. I think absolutely, but I think also To kind of keep in mind that the platform isn't necessarily just one product And it isn't necessarily something that you will need to adopt It could be you know, you at most organizations already have any number of tools that are being used by by our engineering Or engineering teams and it's like how can we perhaps streamline those processes and look at the friction already? within the existing platform or the existing toolset and Try and make that perhaps a more cohesive whole and absolutely as Victor said is if you have The Your end goal or your end destination in mind of what you would like it to be that certainly helps And it's something that's been mentioned in a number of the talks today That whole platform is a product approach and it is about having a roadmap and engaging with Your end customers to our engineers as to where we want to go in the direction that we want to take platform in I think those are really great points If I could just summarize what it meant for us and what it means for me is like is the supply and the demand right the Supply is the engineering muscle that you have around you that you can have access to and then the demand is of course Those who will use the platform and I think when you choose is the gap that you need to bridge, right? So like if you can build everything yourself, that's awesome And you can maintain yourself most likely that's not gonna be the case in most of the times But if that that's the ideal scenario, right? But if you cannot do that then that's where you kind of start shopping around and see what is There in open source. What is there in a closed horse and so on and I really like that you mentioned maintenance, right? Because building something new and shiny. It's really it's really great and it's not that hard usually It's really the maintenance and like those years Long long years where the platform needs to survive and evolve is that where I really see the biggest challenges And that's actually really ties into the Open source and the vendor that journey that the product goes so Yeah, fantastic So there is this concept that you mentioned that is kind of invalidating my question, which is great that There and it kind of came out in different questions on how there is no tabula rasa Like there is no like an empty state when we start being a platform There's already prior artists things that we are using that we it's they're just there So then when you build a platform you actually have to think into account what is available and what people are working with Before you can start like shaking things up Even if you just you have a clear goal of where you want to get that adoption path It's gonna require some some efforts more than even greater than the technical side. So Let's then they take them like case by case Because that way we can get some more insight on how you think about this problem You mentioned the ink on that if you can find something that it's like ready to buy and that can be useful to you You can just buy it because it's like a no-brainer that you wouldn't want to invest Engineering time on building something that's already feeding your your purposes. So I would like to know more about what kind of criteria I would make that kind of ideal Bend or ideal option for for your platform when you decide well, oh, I have to buy these components If there's something where the vendor does will take us years of engineering to achieve There's something we will buy and To buy to to in order to invest on something like I had mentioned that you need to know what you want First so at least out and the requirements What's must have what's nice to have Eccentral and then you you can start really considering isn't this something we can build our self and anything we can adopt Maybe there's something off on the shelf and they already miles away Then everyone else so there's no need absolutely no need to try to mimic in the behavior and Not to measure about the the maintenance and the further bug fixes enhancement Eccentral it's gonna cause fortune. So yes if some something someone else does much better than us Adopted they already start start Accelerate your own development. You can use it to I don't know better serve your users Like if someone does tracing so much better than yours, why would you spend so much engineering hours to write a tracing program? This doesn't make sense Maybe and I absolutely agree like if you can buy something off the shelf That's great But they just maybe just put a little bit spice into the discussion is like the vendor looking right is like the thing that we all Know and and I think it's it's unavoidable right most of the times you probably end up in some fashion in it But I think one thing that I personally this is purely my opinion What I would be looking out for is is there a monopoly in that field, right? So if some vendor or someone is actually monopolizing it There's a higher chance that the terms and conditions will change throughout the time, right? So that's something to be careful. It's much better to be in a space, which is you know kind of full of Competition and you know innovation and everybody's just trying to do better. I think that's much safer I feel much more calm getting into that As opposed to like you enter some really niche area and then you have like really specific vendors kind of taking over So so that's what I would be kind of careful about Maybe I can share two standards so that we follow in Saksa Bank when we buy something so nice Definitely open standards. So the vendor has to follow open standards It will enable us to switch later another thing is data Probability you should be able to control your own data if your data is elsewhere You are already starting at the wrong path. So take a control of your data. Make sure you can get it out later I think chain Alice is a little bit smaller than Saksa Bank and like So for us, I mean costs comes into the equation and certainly something that we would look at and would feed into the evaluation process but along with Our platform team is you know with seven engineers So we have to look at the skill set that we have within our engineering team But also our appetite. Do we want you know if we're adopting a particular, you know, kind of tool of particular technology What does that? demand on our skill set and Kind of engineers look like for the next few years and so those are all things as a much smaller team I think we have to also take into account I've seen organizations of all sizes even big financial Organizations at a later point regretting buying because it can just become so expensive. For example Observability is something that is very necessary, but and many of you might have face this situation that the bills can really add up quickly and Even if it's it's an issue even for big organizations I think it's something to always consider and as platform engineers We should try to leverage the standards provided by the community so if there is ever the need to roll back a decision we can and Are not locked into to a specific bend or I think that's something Yeah, a lesson learned the hard way Yeah, that's an absolutely good point that it's been mentioned Of the vendor locking that it's a very big risk And as you mentioned it can be contractual It can be because the technology might be outpaced and then we have to migrate to something else or the vendor Disappears or is a choir and tries to buy or sell us different licenses, which is not taken from real life recently But Yes, it's a risk. So then another option is to go for standards or open source Technologies or projects that are available. However, as we have seen open source is also not the miraculous cure It usually requires effort to set up and it also is quite shifting because it's supported by companies and these companies may like Like close shop or they might just switch their priorities So that is does it's not an entire guarantee that the open source is gonna keep us going for the many years I would have this platform to run so so how do you assess open source options when you are Going to say like oh this exists in the open source space How do I say oh, I'm gonna choose this project over the other or yeah, how do you know about that? Github stars Well first thing that I maybe look and who is behind it trying to see that it has a roadmap ahead It's going to be supported What else I agree, I mean I would look at the activity within a community, you know how you know how fast it's developing What the contributions look like? But also I would hope if it's something that we're going to be putting into our production Environments for our engineers. I would look for a certain level of maturity and stability With the tool So generally it's it's unlikely to be you know We might kind of experiment with the kind of newer projects But it would take a lot for us to kind of push that but out and roll that out on a wider basis Yeah, and I totally agree that you know open source projects will not survive just purely on a good will So most likely there's going to be some companies kind of contributing and maybe will be the primary Contributors and that's where another kind of you know question mark comes in like okay How is that future of that open source project look like right? What what is really kind of how's going to look like in five years and son does another risk that I would be always taking to account right? Because there's a high probability that that's maybe gonna become close horse, right? We've seen that times and times again, right? I'm don't want to jump into another subject, but it's already kind of pre-facing that so it's really hard I would say it's really hard because Ideally you would want this really sunshine scenario where everybody's just contributing and everything's open source That's not really sustainable and then on the other hand you have like maybe Enterprises will just take it over and it's gonna be not really an open source anymore So striking that balance is really really hard I don't have like you know silver bullet for it But it's just something that when you assess you really need to kind of look into those both sides. I Think we touch a bone almost like all in the area active Community and then it should have a road map. Maybe also good documentation, etc We also look at who's behind who are the big players behind the open source projects I actually much rather of it from many of our assessment is having big companies behind them They might influence some of the road maps However, because they are big so they can support them financially. It's a sustainable way to going forward And one more thing we'll think about is the license. So I mean we have heard enough about Open tofu open bow, etc. I'm glad all the Chinese cuisine got pushed forward to the world. However, I Hope you may get introduced in different contexts. Then this in then this one. So We we feel much safer if it's CNCF project We know that a nice environment for it to grow and to to navigate the government it So I think yeah, there are many things we should consider before adopting any open source projects And one more thing is maybe You have to think if you ever need to contribute. I hope many Are considering to contribute also to the community You have to think also in terms of the people that will run this internally. Do you have? enough people that you can source to participate or grow in these teams and If you work only in a specific city is are there enough developers that you could hire at some point You also have these limitations with With open source projects that You need to see how you will maintain it and make sure that you can That you have the in-house expertise or that you can recruit for it if you don't Yeah, that's a very good point as well because As it's been mentioned Are we not only want to adopt this now? But we want to make this something that's going to survive for years and that that means that the open source project Is gonna last and that includes having some kind of Sustainability back or somebody who can back it enough For it to develop on its own or that there's many criteria We are very fortunate that CNCF has a very mature way of making projects go Forward and become very stable. We have seen And yeah, that's great And then with that that says us already into the conversation of day two operations Which is something that is less talked about because we are most of us are in day one But some people have already already running the platform and I have to maintain it and that is a big Area of interest to me like once you you have your initial setup or okay We bought this part we build this section and we adopted these frameworks Like a year from now two years from now. How do you run the maintenance of this whole platform? So it keeps running in General or if you have to replace something or yeah, how do you go about it? We have reviewed our platform a few times now And then not to mention how many components will swap in and out and I see that as a healthy way Going forward and so nothing stands still and you need to have a capable team So people come into play here. You need to as organization You really need to spend time and effort to grow your employees Send them to KubeCon for example and send them to various meetup group to grow them So your team can constantly get exposed to new way of working new way of thinking a new way of development To keep your components alive. Otherwise. I mean, we don't want to be still like a mainframe So we have to I see that as a healthy way that they keep changing and keep evolving your components your platform I think in what is also sorry what is also really really important be it Vendor or be it open source is being very very close to the you know to the product So if it's vendor then for example, I had created hundreds of different tickets with different vendors before I had a very long Conversations about different bugs and fixes and you know feature requests because if you don't get involved You know that kind of will just take its own way, right? And that maybe will not be the best for your organization or the team or or so on and the same goes for open source Right like I see often people you know like complain about open source powers But we sometimes forget is like it's it's it comes for free, right? So be involved try to Contribute try to be part of the ecosystem and that's just at least some way to ensure the better future for the platform Now when it comes to operations itself like of course We have a lot of good practices and I'm not gonna dive deep into that But but for the open source and vendor would say keep close yourself to the ecosystem. That's super important I think I'm maybe thinking about your question in a slightly different way So I'm thinking about our platform and how we evolve it and how we look at the tools and what needs to be changed and what needs to be updated and for for us at chain artists, it's very much about Trying to get feedback from our engineers. How are people using it? How are they finding? How easy is it how yeah, how easy is it for them to actually be able to achieve what they want to achieve And it may well be that there are certain component parts of our platform that are no longer fit for purpose or that there are Other better products that we need to evaluate and really that's our job as platform We need to continuously strive to try and hopefully keep on top of what technologies available and how we can use it to Better support the engineers Yeah, I think that's a very interesting way of thinking about it and there's the Let's call it the health of the consumer side the developers that are using your platform and also the health of the producer side the developers that are Running and maintaining your platform and you have to look at both your garden our organization has to take care of both and Make sure that the teams are well staffed that you have plans for continuity keep training people keep Don't depend on a single person that if they leave the the project is on a difficult situation All these yeah, those are all fantastic points I and I think it's it's just rather than the perspective of even what I was thinking about with this question because you Started thinking home with the idea of it's not wrong something didn't go wrong if you have to change your platform It's normal and it's healthy that you have to replace it especially in this place that is just emerging It's natural that a year from now We're gonna have a better product or a better technology or better vendor or even ourselves We have learned that this is something that we don't need based on either from the feedback of the users Or because it's becoming too hard or too expensive and and then we have the perspective of having that like Relationship with the people that we are providing us with services And I think that's really crucial to be close to your vendor or be close to the open source community That's building the project that you're developing with so that's a really valuable and insights And then what you mentioned is getting ahead As a conversation, but it's the point that we're gonna touch on now of how do you evolve your platform? How do you take it because you in you we started the conversation with you saying that you have to have these goals and Then we decided to build it we put something together We are maintaining it, but it's very unlikely that we are gonna achieve our goals in the first iteration We need to make this platform evolve towards something. How do you? Plan for evolution. How do you make this intentional change over your different components and the sources that they have? Yeah If I must list two most important thing Well, we always work towards you. Let's make the platform more accessible and the more disaster recovery ready So disaster happens which we yeah, we have encountered So how do we bring the platform back up as fast as possible? This is something we constantly work towards to second thing is how to make in the platform more accessible to our employees our developers Because in the beginning you focus on functionality you get the functionality out so they have this thing they can utilize But it might not be as pretty you might never have might have to access via CLI or API etc You want to make it as simple as possible for everyone? So it may don't need to be an expert They don't need to be a super user in order to use your functionality provided by the platform Yeah, for me when we start changing a platform we Try to take a cautious approach releasing it and testing it with some of our users before Taking some metrics and then seeing if it's the right way to go and in that case And Be customer centric be customer oriented the way not to do it is to introduce and these expensive migrations in terms of time where it will take the time of all the teams to change to your new system and Create a lot of frustration in the process. That's the way not to do it If I could add and I'm not the first one to say this but like you know eat your own dog food So when it comes to the to the users, right? I had a kind of fun a funny anecdote where we were preaching for the spec first approach for the apis and then some one of The users came and said, you know this API that you have exposed is actually code first So why are you even preaching that right? So that was really really kind of Kind of a funny moment and I think it's really important So you need to walk the way with your users You need to put yourself in the shoes of a users and that journey might you might start that here And then it's gonna end up completely different place and I seen that time and time again So I think it's super important. Yeah, eat your own dog food try your project products and also walk together with your users And I'm gonna kind of use that buzzword again or the buzz phrase and the platform is a product and it's something that we really try and work towards It's not always easy. We don't have a product owner a product manager for platform So all of our engineers all of us have to embody a little bit of product in our interactions with engineers, so it means that we have to talk to engineers and find out what the issues are and then try and build that into our roadmap for how we want our platform to evolve and Kind of balance that up with and planned features and plan developments that take the engineering organization Forward so that may well be kind of security guide guard rails As well as you know suggested features from our engineering teams as well Right Yeah, you have something. Yeah, I actually had a common to victors migration. We just go through a huge migration from this pipeline pushed approach to github approach and There was a hell of a migration, but it was totally worth it. So if you guys are not github's ready, please use flux or Argo, they It will pay off in the longer run. So sometimes migration is necessary. Sometimes migrations can be good Yeah, there are ways as a platform team that you can try and make that an easier process And it may well be that you know, we could invest time to make it, you know A few click buttons rather than a few weeks of hell Definitely Yeah, fantastic. I yeah, I think that's I don't know how we're doing with time, but Okay, then thank you so much and if anybody has any questions Hello, so you all talked about value and the value of the platform will bring to the business Can you give me some ideas on how you would measure that? Or what it means what value it means to you if I yeah The typical value is we compare our Platform speed deliver speed tune or to ticketing system so the existing process Let's say if you want to have a database with access Well, your whole group need to have access to certain database with in certain sites certain spec How fast can they order via your platform to on the traditional? Ticketing system like service now, etc. So if we can prove we can do it faster I will say next value right there Dora metrics for example, you can use Dora metrics to validate that your organization is moving faster and safer What else customer satisfaction I like a net promoter score you survey people and You have promoters and detractors and then you Make a sum there and if it's positive, you're doing good and the higher the better Absolutely, I was gonna say pretty much where you said and so it's like that mixture of quantitative metrics that you can kind of imply or you can work out from the data and Various, you know peer pull requests, etc. As well as qualitative information And I guess it's soft decide, you know are How do our engineers feel about the platform? Where do they feel the bottlenecks are and sometimes that can surface issues that you just can't find with the data and I Agree fully with all the points But I just want to kind of put a little bit that actually measuring is very very hard And I'm not gonna name the consultancies that were measuring developer productivity recently I think that's really was a hot topic, right? And I actually think I tried really hard to measure I worked a lot with observability of our platforms and I really didn't find the silver bullet How do you really measure that so other than concrete cases, right? Like somebody did a project this way and now they did this way and then they come back to you say this is great I don't disagree Dora metrics are awesome But we really when we look into data would need to be very careful what we're looking at because there's always this bias that we have in data Hi so You're sort of running like a Internal product start up within your orgs. How do you market it and sell it within the orgs? Do you like use the carrot method or do you use a stick method and we share some examples? For me, it's literally everything whatever works and sometimes it's it depends on what initiative you're pushing out You've got to try and use a whole set of tools and sometimes it is literally For some projects, you know, I've literally had calls with every single team and invested the time and actually I have to explain Why they need to make time in their spirits to dedicate to this and why it's important So yeah everything I Think education is really really important. We do a lot of training online training or workshop, etc It's like you need to write really rather campaign for for your setup to sell your your product It's actually like a better a platform as a product. Why is your product a better name other products? And I would say we always start with a carrot, right? But then at some point You might need a stick but to be honest like communication is the key I had some experiences before that when you go and really talk with the users you find out that actually They don't adopt Because the platform is bad or it's wrong, but just because their needs were not heard early enough And then you really need to listen carefully. What is exactly blocking them from adopting it? So most of the time it's not going to be because people just don't want to it's just because it's not making their life easier And that's really the key why we need platforms, right? When you're introducing something great that is a carrot in itself I think if you really need a big stick you have to think why but sometimes it's just the nature of their problem Sometimes it is necessary In terms of these big stick, I would probably Translate that as leadership sponsorship. It really helps I found you know if you can get you know, CTO say, oh, do you know what? This is a really good idea. Suddenly everyone's listening Hi, is there any thought or any requisition from the management in your organizations? About how the platforms can make the organization more resource efficient more sustainable, maybe I think so, sorry Just to understand sustainability you mean environmental sustainability or Yes, yes, absolutely. So this is really really important topic in our organization and and the efforts are constant We really if we look really from engineering perspective, we actually start measuring like okay How much you know resources will use what we run our applications. We we done some efforts moving to different instance types We've done different Investigations like even programming languages how much if you run it in this program language and the different one This is what we as engineers can do. There's much larger efforts around it But I think all of us can chip in at least a little bit. I Can't agree more ESG is such an important topic. Thanks for bringing up We commit to to improve How how can we reduce emission and how can we for example, just why do we run in run things in containers? Why do any Kubernetes to increase the density of an application we can run in in each server? Why do we care about which language we write? It's just because of my efficient than any other. There are many things engineers can do to improve Environment improve our CO2 emissions. I think everybody need to keep that in mind And try to build a platform to to go towards to that Templating for example, if you can template the way how you build Application put all the logic in then people doesn't need to write everything from scratch Small things like that can can go a long way. I Think we've got time for one last question Hello first of all And There is a big difference between the users of a platform and the users of another kind of products I mean the users of a platform also have a technical knowledge So it is possible to use that knowledge in the into getting a better tool for for for a world organization In this sense an inner source strategy can be Can come handy. So how do you incentive this kind of strategies? inner source as in Getting people from the developer teams to build for your That's a great question First you have to encourage I think openness in your Organization and you have to make sure that it is easy to contribute to your systems and you have to communicate this as well and I think you can notice people that are interested in the technologies when you are interacting with them on your daily work and you can approach them if your organization allows this Absolutely and it's all saying at chain artists particularly with kind of our platform is really important given that there are only seven platform engineers serving the engineering department or engineering and We do that exactly by a kind of encouraging Contributions to to the platform and you know as Victor mentioned, you know, you see the same generally the same feature engineers will contribute again and again So we have like you know You know kind of really keen engineers that do want to get involved but it is also a little bit as we mentioned earlier about I Making our repositories and are the platform as accessible to Feature engineers, so it does mean if we're saying that everyone has to have a great read me and You know great documentation around it We have to provide that for our repositories in order that it is accessible to allow the engineering the wider engineering and Division to contribute Okay, but just wanted I agree with all of it and I would say we also need to champion the inner sourcing So if somebody we had cases where somebody has hey I have implemented this tool for your platform and we don't just ignore it Why like we actually bring it up and say hey look at this. This is really awesome like everybody come on Let's do it, right? So I think it's also important to just promote each other's work and not be too selfish, right? I think we all care about our work the most But I think it's also important that we promote actually each other's work as well Thank you so much Yeah, enjoy the rest of the conference