We're not going to beat AWS on their units of economics. But if we create a new way of running compute that is significantly more efficient, that does become a game that makes sense to play.

I'm joined today by Rita, who is the VP of developers and AI at Cloudflare.

We do have leaders at the company who are students of Clay Christensen. The idea of disruption, I think, very much is in the core DNA of the company. We really try to hire people that have all of these traits. We built a developer platform and we use it to ship.

You just launched containers

which is We did.

Pretty big. Does it feel like that's like the is it like joined up now? Because it feels like that was like the kind of at least like one of the crowning pieces that was always like, now it feels like you can kind of do everything on Cloudflare with containers?

In a way, yes. I know that that's something that users have wanted us to release for a very long time. I would say, you know, it it's it's an interesting one in that we've been having the debate internally of whether or not that's a product we release to our customers for quite a few years now, actually. In some ways, we've had the product built out, and we've been running it in production for for quite some time now as well. And so it's really been a question of, yeah, do we put it directly into customers' hands, or is it something that we expose through, you know, if you use workers' builds, you've used our container platform.

If you've used browser isolation, you've used our container platform. And ultimately, where we landed was, you know, there's always going to be a long tail of packages, libraries, things that people need to run somewhere. We have a bunch of these internally too. Again, we have use cases like workers, like workers' bills that we ultimately needed containers in order to be able to run, and, we do dog food for that reason, and so if we find ourselves in a situation of we need something, we generally think of, you know, how do we expose this to customers because they probably need it too. Mhmm.

So it it does feel like a piece of the puzzle that kind of completes the platform in in a a way, way, but but I think what you've been able to do on the platform even without it for such a long time has, you know I I'm both very excited about it and I also don't wanna overstate its importance just because there is so much hype and this is not to say that I'm not excited about it. I'm very excited about it, but it's a balance.

Yeah. You have to kinda do it the Cloudflare way in the sense with the workers whereas like, this is like, you can just do it however you want.

Yeah. That that is very much how how we thought about it. So to give you some context on why we went back and forth on, you know, whether we release this to users over the years too, The reason we even entered the developer platform space, the cloud space, the compute space is really started with the release of workers. And workers came out of a constraint that we had where, you know, the amazing thing about constraints is constraints breed innovation. Yeah.

And we really had to allow developers to program the CDN. And if you're going to program something like a CDN where the value of it comes from it being really fast, it means that it has to be really fast and really efficient. Right? You can't spin up a container that's going to have a ten second cold start on every single request. That's nonsensical.

And when when we stumbled across workers, which was obviously intended for one purpose, but we we realized it was powerful. The reason we were then like, oh, actually, it makes sense to build an entire developer platform is because it was a disruptive technology that allowed us to provide compute in a way that was orders of magnitude more efficient. We entered it from this way. If we'd entered it from the other way, so let's say we woke up one day and said, we want to compete with the giants, right, that have been around for over decades now and have specialized in running centralized data centers and running really, really massive data centers that are designed to run many, many VMs. We're not going to beat AWS or GCP or the other providers on, you know, their units of economics.

That that that doesn't make sense for us. But if we create a new way of running compute that is significantly more efficient, all of a sudden that does become a game that makes sense to play.

Yeah. So you have some like software a software advantage, not just like because you couldn't have like a hardware advantage in the sense of like as many having as many data centers and like the scale of it that they have.

Exactly. That's such an important part of why do this. And disruptive technologies do disrupt over the long term because they are more efficient, they open up more doors, they make accessible a technology that would otherwise be cost prohibitive in many ways to to customers. Right? And the way that we see this show up is not just in terms of raw cost, but also the the thing that was so interesting to us was the ability to deliver on the serverless promise to developers or actually really the cloud promise to developers of not having to worry about servers or infrastructure.

The problem with VMs and containers is that ultimately, no matter how much you try to abstract them away, it's a leaky abstraction. As a developer, if you're trying to scale them up, you have to worry about regions, you have to worry about concurrency, you have to worry about cold starts. We know of many customers who, yeah, maybe they're using a serverless function, but they have a whole Rube Goldberg machine that's set up around keeping it warm. Right? With with workers, we're able to provide something where you actually didn't have to do any of that all.

And I think that that opens things up to so many more developers. And it it's interesting now having the developer platform in this kind of cross section with where we are with AI too. So I feel like we opened up the door to so many more developers on the infrastructure side, and then AI is opening up the door to so many more developers and just the, like, writing code side that I think the two together will be really, really magical. But, yeah, to to go back to workers versus containers, that that's the origin story of where we started. And, yeah, it's I feel like now that we've proven that you can actually build a whole developer platform on top of workers and on top of isolates, containers was just the piece that we could slot into that that creates the picture.

Scaling DevTools is sponsored by WorkOS. If things start going well, some of your customers are gonna start asking for enterprise features. Things like audit trails, SSO, SCIM provisioning, role based access control. These things are hard to build, and you could get stuck spending all your time doing that instead of actually making a great dev tool. That's why WorkOS exists.

They help you with all of those enterprise features, and they're trusted by OpenAI, Vercel, and Perplexity. And if you use them for user management, you get your first million, yes, million, monthly active users for free.

I honestly don't know any dev tools that have a million monthly active users apart from GitHub maybe.

So that'll get you pretty far. Here's what Kyle from Depot has to say about WorkOS.

We use WorkOS to effectively add all of the SSO and SCIM to Depot. It's single handedly like one of the best developer experiences I've ever seen for what is, like, a super painful problem if you were to go and try to roll that yourself. So for us, we can effectively offer SSO and SCIM, and it's, like, two clicks of a button, and we don't ever have to think about it. It's, like, one of the best features that we can add to Depot. It's super affordable, which effectively allows us to like break the SSO tax joke.

And essentially say like you can have SSO and SCIM as like an add on onto your monthly plan. Like, it's no problem. So it really allows smaller startups to essentially offer, like, that enterprise feature without a huge engineering investment behind it. Like, it's literally we can just use a tool behind the scenes, and our life is exponentially easier.

Like, how are you able to kind of like think about this? Like because I feel like most people don't have this like strategy that Cloudflare have. I don't know that like like Swix has written about this really good article about like how you're playing Go, how you're like surrounding the big players. And then I think Strathecari have also ran about Cloudflare and how disruptive it is and like how I think he was like referencing Clayton Christensen and, you know, all that sort of stuff. How does that like actually manifest behind the scenes this kind of like strat strategic thinking?

I mean, we we do have leaders at the company who are literally students of Clay Christensen's, including our founders, Matthew and Michelle, and James Ulworth, who's our head of innovation and actually co wrote one of his books with him. So so the idea of disruption, I think, very much is in the core DNA of the company and permeates it throughout. And if you think back to even some of the original products, we always looked at it from the lens of disruption and taking services that were previously only available to enterprises, making them commoditized and available to everyone and and disrupting things from that standpoint. In terms of, you know, how the mechanics of of it work in the day to day, I think we are generally a very long term focused company and in everything that we do, we try to ask ourselves what's our what's our competitive advantage? What's, you know, what's the unique thing that we can bring to the table?

Especially when it comes to our developer platform. And so that's where a lot of these decisions come from. It's like, you know, again, we don't want to generally some some of our products do come from, you know, if if you think about all of the pieces that a developer needs in order to build an application, obviously, want to provide as many of the primitives as possible. But at the same time, whenever we do bring a new product to market, we do try to think about how do we build it in a way that, yeah, that that allows us to compete over the long term, not just over the short term.

I guess that's kind of coming back to containers was like, maybe that was why you were delaying was like, what's our competitive advantage here? Whereas like with workers, it was clear where the competitive advantage was.

Exactly. Exactly. And since since I joined the company, the three key values and sorry. I'm thinking because I don't wanna get this wrong. But the the the three values, honestly, sometimes when companies talk about values, I roll my eyes a little bit.

You do feel it so much when when you interact with people here and it's curiosity, it's transparency. And then the third one is around oh my god. I'm like blanking on the word itself, but not the principal. Oh, principal. Oh, nice.

But I I was not intending to make that joke. But but I think, you know, the this applies when it comes to if you look at, you know, no one likes to have an outage. Obviously, those are not the fun moments for us. But if you look at, you know, our responses over the years, the one we have had incidents, we try to respond to them as quickly and as transparently as possible, and that's something that we're very, very principled and disciplined around. Similar thing with you know, ultimately, I think, again, it's very much in the DNA of the company to be in the business of trust even if you think about just sending traffic through us and letting us decrypt it in order to, you know, provide WAF mitigation, in order for someone to put that trust into us, that means that we're not going to take certain actions that maybe someone else would see as profitable in the short term.

But in the long term, that doesn't allow us to grow a sustainable business.

Yeah. That that does fit with what I've kind of seen. And I was just speaking to a startup today that are pretty much building on top of Cloudflare as like as an explicit thing where like they're just trusting you guys to just handle all the infra stuff and they're not gonna worry about it. So that that does make a lot of sense that you've been doing that. And I guess it's stuff like r two fees and people sometimes like how are you making money off There's some like a lot of the DNS staff.

It's like, I don't think I've paid money for that.

Having a free tier. Some people ask, you know, why you guys do this? How are you gonna do it? Is it gonna go away? It will never go away.

That's awesome. I get and then so kind of well, segueing. You have a very active Twitter kind of team. And so online. It's like well, it feels like everyone, especially London, like all of the people that I'm like, everybody I saw like Kat's just joined.

Right? Kate's just joined. And so, like, I felt like it's like everyone's in the team now. It's like in London, it's like like Sunil and you essentially have a lot of very talented people and they're very, as you said, online. It's almost like DevRel's, but they these are the people that are actually building the products.

I think we just try to hire people that really deeply care about what they do and that comes with you know, if you really care about what you're doing, you want you care about what people are saying about it, first of all, so you're always monitoring for, you know, is my all of PMs are and and it's not just Twitter. We're in the we have a developer Discord as well. It's a great source of feedback. So we try to yeah. You know, first of all, monitor things for feedback and be really responsive to people.

But similarly, I view you know, I was talking about curiosity being one of the values. Right? And I I think similarly, as much as I wish I was less online than I am, especially right now with AI and with things moving so so quickly where every day there's a new model being released, there's a new product being released, you kind of have to keep your finger on the pulse and understand what's in the zeitgeist in order to build a relevant product and make sure that you are keeping up with what's top of mind for developers. So it's not that we hire for that in particular, but I think it's not entirely a coincidence either. Yeah.

But we also have a lot of great people, you know, it it takes a village to build these products and there are a lot of amazing people that are not on Twitter and still do incredible work.

Yeah. Yeah. Yeah. I I'm I'm very sure of that. Yeah.

It feels like there's been from the outside, and I've seen it's not I've seen people write this as like Cloudflare are like just shipping like crazy these days. It feels like you're going faster. Like it's not which is usually I felt like it's the opposite. It's like you kinda get bigger and you're you're the people I felt like people don't I didn't maybe even realize it's like a the the market cap today of Cloudflare is like 60,000,000,000 or something. It's like, this is a big company.

This is not like a startup, but you're still pumping out a lot of stuff. How how do you think you've been able to do that?

We're just churning out AI slop.

So chat, GPT. Well,

no. I think we've, you know, our number one secret weapon is we've gotten we built a developer platform and we use it to ship. I mean, that that is my promise to developers. That's the thing that we go up and talk about is when people ask why should I use Cloudflare over insert hyperscaler here, the answer is we allow you to move faster, and that that's the way that we've enabled ourselves to move faster as we built the developer platform that we needed in order to do that. So that's certainly a big piece of it.

The second half of it is, I think, again, there's very much a culture of that that encourages moving quickly and, you know, there's there's this idea that I think when you're a startup you have and as you grow up, you kind of lose it because the expectations grow where, you know, if you're not a little embarrassed by what you've you're shipping, you've you've waited too long. And this is a lesson that we've learned over and over where, actually actually, the the faster you put something into customers' hands, actually, the better product you'll build and the faster you'll be able to go in some ways because nothing ends up being as motivating as receiving a constant stream of feedback from customers that are telling you like, oh, I wish I had this feature. Oh, you know, I wish this was faster. It makes you realize how much you know, what a big backlog you need to get to, but also it's very motivating to then put the next thing in customers' hands and see how excited people are, see what they build on it, and that gets you to, you know, go build the next thing from there.

I think where a lot of teams get stuck is, you know, I think it's easy to get stuck in the early ideation phase of, you know, what does this thing look like that I'm supposed to build or how should I build it? And we very much, to that extent, encourage prototyping and MVPs, and we do weekly demos where it's like, you know, like, let's let's just put something out there, get a feel for it, and and iterate on it week over week. And then, yeah, I think the other place that people get stuck is in trying to perfect things, and you're you're never gonna get something perfect. I I think actually the risk of the longer you withhold something to put it into customers' hands, the risk of getting something wrong actually kind of goes up.

Getting it out there fast makes total sense. And especially if it's like in the SaaS context. But I felt like it's way harder when you're like a you know, this is an infrastructure product. Like, if you launch something new, it's like people are like relying on it. It's like, how how have you managed to like kinda keep that like mindset while knowing it's like, but it can't it can't like break or it can't like, you know?

So for that, we have different guidelines for different products and their levels of maturity. Right? So the workers' runtime, for example, we've had to, first of all, think very deeply and come up with instrumentation that would allow us to ship faster. There was a point where we were kind of stuck in being able to release changes or change APIs, and we realized what we needed was feature flagging system in order to allow users to opt in because we didn't because breaking changes would be we would just break entire websites that are on the web, which is is unacceptable. And so we ended up building a feature flagging system.

And even with that, something like the Workers Runtime, we're going to be a lot more conservative about releasing it. And again, this is where we build things into our platform to also allow us to you know, we will allow releases to we'll roll them out to Canary first to a certain percentage of users and allow them to bake for a little bit and see, you know, over the small percentage of traffic, are we seeing any changes and roll them out much more gradually, any products that are built on Workers. Workers has this gradual rollouts concept, so a lot of our teams release in this way too. And so you kind of have to use your judgment of how conservative versus, you know, a product that when it comes to releasing a new product, if it has zero users today, you can go really, really fast. Right?

Yeah. And so I I think having that level of granularity and judgment is actually really important because sometimes people take things very dogmatically like, oh, well, you know, this is how we release things and it's like, no. I actually don't need you to bake something in Canary for forty eight hours. If it has your users, you will learn nothing.

Yeah. Yeah. Yeah. Yeah. So it's like you're yeah.

You're not using the same process for like something someone shipped at the weekend and

Exactly. Exactly. So something's gonna have a different life cycle if it's still pre customer. Similarly, this is why we have, you know, alphas and betas where

we

like to give ourselves a little bit of wiggle room in terms of, you know, we might still change certain APIs even even though, honestly, normally, once we put things into customers' hands, we we assume that we won't be able to change things very drastically because no matter how many times you say something is a beta, someone out there is relying on it and is never going to update their code.

Yeah. This is true. I know that Sunil's working on the agent stuff. Is that, like, kind of how you see you're seeing, like, the future of things as everything's I mean, probably wouldn't be a very contrarian take to say like, a it's gonna be agents and like Cloudflare's gonna be the best place to run agents. Is that is that kind of the plan?

That's that's kind of the plan. I wish I had my agents hat that that we made. But, yeah, I think we kind of inadvertently built the best platform for building agents, and I really truly do believe that the more the more we're seeing agents being built, both in terms of just the primitives that you need in order to build an agent from voice operated agents, which we have our real time products for all the way to, yeah, chat powered agents. And agents do have such distinct usage patterns in terms of their dependency on long running LLMs to execute and their needs, you know, patterns like human in the loop that require you actually the flexibility in in the pricing model, really, in order for, you know, agents to even be feasible for you to be able to run hundreds, thousands, millions of agents in parallel, wait paying for all of that idle time becomes unfeasible really, really quickly. And so, yeah, I I'm I'm very bullish on, first of all, just agents in general and especially on developers building them on Cloudflare.

Very cool. Very cool. So something I wanted to ask you as well is about like the feedback mechanisms that you have. Because it seems like you're quite responsive. Especially, I mean, I think you're responsive anyway compared to like the other companies that are on the Twitter and stuff.

But, I mean, I guess you're, like, halfway between, like, the startups and the hyperscalers or something. But I think you're definitely acting a lot more like the startups and, like, feedback. So I can't wait to ask how, like, how do you actually, like, get what people say back into the into the sausage?

You know, first of all, in terms of the the gradient between the hyperscalers and the startups, if you compare, you know, the number of people that work at, say, AWS, Like, Cloudflare is we're 5,000 people. It's big. It's much bigger than a startup, but it's not

How how do you think AWS have, actually? I just I don't even have a gauge

on that. I I think it's in the order of magnitude of tens of thousands.

Okay. Okay. Okay. Okay. Okay.

Yeah. So 10 at least 10 x bigger.

Yeah. Yeah. Big difference. I think, you know, I I think the cool thing actually about working on on a developer product is the extent to which you're able to be connected to your customers. And so we incurred you know, I don't have to receive the feedback directly and then funnel it all the way to, you know, a PM to an engineer.

A lot of our PMs and engineers are able to seek out this feedback themselves. And so it's not this long chain of, you know, getting that feedback back to someone. Again, that's that's something that we just very much encourage and not because, you know, it's public, but because the feedback ends up being really valuable both in terms of long term feedback of things that people are you know, if if I see several times that people are have the same question, are are struggling with the same thing, it's very telling of we should update the docs or we should update the product or the onboarding because, obviously, this thing is not very clear. Sometimes it's also very good signal for, you know, if I see that someone complained that, I I don't know, that they're unable to update their worker or their worker subscription more than five times in the past hour. Oh, like, that might actually be an incident.

So the I I think staying connected to customers in that way, I I don't see that being something that we lose as we grow.

It seems like it's like team is basically the key ingredient because I guess you couldn't do that if people were if you weren't hiring really good people that could kind of do a bit of everything, to customers cared about the problem and

Exactly. And and part of it is PM's job too. You know, I I think that is something that's harder to expect out of every engineer, but for the PMs at least, which are literally the job of a PM is to be the voice of the customer and to bring the voice of the customer into engineering conversations. The customer could be in customer meetings. The customer could be at conferences.

The customer could be on Discord. The customer could be on Twitter. And so it's ultimately your job to seek out, you know, those sources of feedback.

How how are like the PMs like, what what does like the teams teams typically look like, the structure?

You know, it depends a little bit often. Actually, it's funny. I there was some discourse on Twitter today about what's the right number of PMs to, you know, products or teams. We generally have a single PM per product, although I'll asterisk that with the fact that Workers is tricky because Workers is a really big product, and it has so many components. There's the runtime.

There's the observability aspect of it. There's the CICD aspect of it. There's it's also the longest running product. And, you know, the the longer you have something around, the more you realize how many features you've launched over the years, and they all need maintenance and updates at some point. So Workers has several PMs that are responsible for different areas.

But generally, yeah, we have we have a product manager that's at least one to one aligned with their engineering team and with their engineering manager. And in certain two things that sometimes will happen. One is there are some products that are going to be at least initially or sometimes for a long time, they're going to have smaller scope, and so it's not uncommon for a PM to oversee more than a single product directly. Or when we go to spin up something new, again, we'll we'll take a PM and we'll even often take, you know, a couple engineers from other teams. We try to have a really, really small kind of tiger team when we do try to spin up a new product because in in order to move fast, actually if you add too many people early on, that will slow you down as well.

And so we try to create this kind of small unit, allow them to run really fast, and if it ends up being successful, if it's growing, then, you know, either that PM becomes full time in that product and we hire someone for for their old role or vice versa, they stay on their product and it's good signal for us that, hey, we need to hire a PM for this role now.

So you just yeah. Keeping it really small ideally

Yeah. As as long

as possible. And so for example, the like with agents, the Mhmm. Product that Sunil is building. Actually, maybe for people haven't come across, could you also share what that is? And and then it'd be great to hear about, like, how that journey's looked to, team and stuff like that.

That that one's that one's a unique one. So so we have a product called Agents SDK. It's it's basically a library you can install if you NPM I agents. You can start running it immediately, and it provides you with a lot of really useful tools for and patterns for building agents. So it comes, you know, if you wanna have an agent's UI, it comes with a really slick UI.

It'll help you manage memory. So if you, know, if you're building an agent, you want to have a long running conversation where it remembers the context and it'll store I mean, under the hood, it's no surprise, a lot of durable objects. So it it'll store all of that in durable object, which turns out to be a very good model for for creating agents. But it it abstracts a lot of that for you and if if you're building MCP servers, we have a lot of libraries and helpers for for that as well. That one was interesting.

How did that one came out of, yeah, us starting to see a lot of people building agents and kind of thinking, you know, what what should we do in this space? This is always interesting discussion, like, okay, we wanna do something. What what is the product? Is it, you know, is it that you go into the dashboard and drag and drop your way through an agent or, you know, kind of crowd UI through it? Is it yeah.

Is it a library? What is it? And in this instance, because it's it's something that we see developers kind of going through and we'd seen a few frameworks also for for building agents, but they seemed a bit heavy handed for us. And so we wanted to provide something that was pretty lightweight and a lot would just allow people to get up and running on our platform quickly. And I think it was over a weekend that myself, Sunil, Matt Silverlock, a few others just kind of sat down and started with the docs.

No Like, okay. What would, you know, what would the docs look like for for something like this? And Sunil wrote out the library and we launched it and, you know That's as

I was

saying before, started to get feedback and realized, okay, this is what people are asking for.

Wait. You guys did this on the weekend. That's what I saw.

That that's how it started. Yeah.

That's well, how did that come about?

Once the mind hive starts going, it's like, okay, let's do this now.

So cool.

I think I mean, it took longer than a weekend to be very clear. But we we set a deadline of like, okay, it let's imagine today is, I don't know, February 10. What if we launch this on March 1? And, you know, this is the plan. Here's everything that we need.

Who owns what?

Wow. That's that's very cool. That's a good sign. Kind of going back to the hiring, but it's like also like the acquiring as well. This seems like you're acquiring a lot of companies with very good people.

I know there's like Boris and Sunil in London, of course. But I know there's a lot more. It seems like a very good strategy from the outside.

People are the lifeblood of the company. People are the most important thing. When I've been in Cloudflare for nine years and when people ask me in interviews and it sounds so cheesy every time, they're like, why are you, you know, why are you around for so long? I'm like, I really like to come into work every day and see these faces. I have literally met some of my closest friends here.

So I I think in terms of hiring, it's something that we care about a lot and recruiting goes all the way to, you know, Dane does a ton of recruiting still who's our CTO, So does Matthew, our founder and CEO. Hiring is something that's taken very seriously. And, yeah, again, we really try to hire people that are curious and humble and empathetic and have all of these traits. In terms of acquiring, I think there's something very interesting and unique about, you know, we were talking about how startups tend to move really fast, and there a lot of the way that we operate is kind of similar to how we would operate a startup, especially when it comes to launching new things. And so startups happen to be a really good source of that mindset and of that DNA of, you know, okay.

Let me optimize for getting something done. And I have finding people that have a really strong vision and a vision that's strong enough to help them, you know, tear through whatever red tape and help them not get stuck and help them carry things over the line. And so it these these small acquisitions have been, you know, a really, really big success for us.

And I guess especially if they're already building like with Cloudflare, it seems like both of those examples, they were like very very in the Cloudflare ecosystem. I know Sunil had already worked. He was like a boomerang.

Sunil is a boomerang. But yeah. I mean, that's the other thing too is the folly of so many acquisitions is the integration piece. And the reason that integration is hard is often, you know, technologies between companies are not compatible. And if someone wasn't built on Cloudflare, trying to get them into Cloudflare would be really, really challenging.

And so having yeah. The more companies build on Cloudflare, the the easier it is.

So perhaps someone could just, like, see where there's a gap in Cloudflare, build their own startup. So adds it to Cloudflare. I mean, slide into your DMs.

100%. If if you're trying to do this, the recipe for success, please please do it. I mean, the other thing too is what we find is if you're building on Cloudflare, you will have a lot of feedback for us. And so we want we want people that like, some of these conversations start out of people continuously tweeting at us and being like, I found this bug. I found that bug.

I found another bug. And I'm like, okay. You you wanna come fix them?

I think it was Sunil saying that was like how he joined or something. Yeah. Am I misremembering? Yeah.

That that was part of it.

Very, very cool. Very cool. Okay. I think we're coming towards the end, I wanted to ask if there's anything that if you had to kinda put your finger on something that Cloudflare does that's quite unusual that you don't see many other startups doing that maybe they should do. Is there anything that kinda comes to mind?

I know that the answer is yes because when people join, things don't make sense to them immediately. I mean, one thing maybe that's actually become commonplace but didn't used to be commonplace is innovation weeks. We've been doing birthday week for many, many years. We noticed that there were things that were sitting at that 90% completion point, and it it was a good forcing function to, you know, go, hey. Okay.

Time to launch these things. And that was one half of it. The sec the actually, the half of it that I just mentioned came later. The bigger part of it was that I think is also perhaps quite unusual about Cloudflare is the obsession with the mission and how mission centered we are. And, yeah, always shipping things for birthday week that were contributing towards the betterment of the Internet.

So one of the first things that we launched as a part of birthday week was free SSL for everyone, wasn't at all at the time. No no one was doing that. It it wasn't at all commonplace. And I think in general, yeah, this idea of taking things that were only previously accessible to large enterprises and commoditizing them over and over again. And even if sometimes that might externally look like, oh, aren't you going to, you know, cannibalize a part of your business because people aren't going to upgrade to get a certificate if it's better for the Internet?

I think it works out to be better for us for the long term as well. So I I think I just named a couple of different things. But, yeah, innovation weeks go way back, and it's funny to see that now a lot of startups do launch weeks. But we yeah. We we'd been doing them for for a long time, and I guess the part of it that comes across as unusual to people sometimes is they can be quite chaotic.

Things come together. Things are both planned for a long time and also come together at the very last minute. And, yeah.

Amazing. Rita, thank you so much for joining. That was so fun. And I wanted to interview you Fei just so really happy that you've had the time to come on.

The pleasure was all mine. It was so much fun talking to you, Jack.

Thanks. And thanks everyone for listening.