I think that would be an amazing video. Steve is a lovely host. I'm very curious because Steve is going to do something that I would not have done, I guess. So, and... Oh, wow. Oh, that looks pretty cool. Alright, so that is the Oxide Rack as it arrives. And then you can see that we've got a built-in ramp there. Oh, right. Oh, it's so smart. Yeah, so this is where you get to all these little bits that are engineered, right?
So those bolts come undone, that small piece comes out, and then the whole rack rolls out. Side panels are shipped, are long with it. These are the side panels here. You've got the front and rear doors. Wow. And this particular... oh, God. Oh, wow. That looks heavy. Welcome to the Software Misadventures podcast. We are your hosts, Ronnock and Guan. As engineers, we are interested in not just the technologies, but the people and the stories behind them.
So on this show, we try to scratch our own edge by sitting down with engineers, founders, and investors to chat about their path, lessons they have learned, and of course, the Misadventures along the way. Take it a big one. Oh, yeah, maybe you'll be fine to start with the office tour. I'm kind of open to prepare. It's like, okay, we're doing an office tour. Let's do it. Context, context. So we started talking about this right before we hit the record word, and thanks, Guan.
And, Ronn, we see this amazing background behind you. It seems like a warehouse, but actually it's a very cool office. We got a glimpse of it last time we spoke. We thought it'd be really cool for our listeners to also get a glimpse of the office. So if you were in mind, can you just briefly show us what the office looks like? Yeah, sure, absolutely. So we've got, we are here in Emeryville, California, and we've got, this is the Oxide Office. Let's see here.
So it's funny, because we were really born right before the pandemic. So we got this space in late 2019 with the idea that we would be doing, you know, we knew we were going to be a hybrid local remote company, but assumed that a lot of stuff was going to be local. And we definitely used the office a bunch, but we used it as you can kind of see as kind of, as a lab, more than almost anything else. So, a very open plan, but then importantly, we've got a lot of hardware. This is our switch.
Yeah, exactly, right? This is our switch. This is our manufacturing station. So we've, not something we've really talked about too much. But we actually wrote all of our own manufacturing software at Oxide. And in addition to everything else, and this is what that manufacturing station looks like at our manufacturer. To be clear, that is not where we manufacture. Oxide. This is not where we manufacture. I know. People think like, oh, this is like the factory.
No, no, no, no, no, no, no, this is not that. This is not that. This big crate here is an Oxide rack. Inside, there's an Oxide rack. Actually, the crate itself has been engineered. So one of the things that was really interesting, and we've learned so much about so much in building this company. But one of the things we learned is all of the engineering that goes into shipping things.
And it is extremely important that when we ship the Oxide rack, that we are able to ship with all the sleds in the rack. So these compute sleds, so here are our compute sleds kind of down here. Actually, I'll give you a overview of the part over here. These compute sleds slide into the rack. And that's a compute sled there. And it will slide into the rack. Our Oxide rack over here.
Nice. And it is very important to us that we are able to, here's our rack that's actually that, this is the, what we call the dog food rack. This is actually a running rack that we run all of our own software on. And it was very important that we were able to ship the sleds in the rack because one of the things that we're really optimizing is the time from when that crate arrives. And the rack is not just powered on, but you've got developers provisioning VMs.
And in order to optimize that time, it's really important that you don't have to debox a bunch of compute sleds. This thing ships with 32 compute sleds and 32 boxes of computers. It's a lot of deboxing you need to do. And you're going to spend it to the day deboxing and disposing of stuff that can't happen in the DC that has to happen in another room. And this just adds time. And so what I think it's very important is that we can, we can ship it with the sleds in.
We can roll the rack in, plug it in and go. But in order to do that, you have to end your own crate. And what see, this is what the third iteration of the crate. I mean, this is, it took us a while to get this, to get this right. But yeah, this has been a, this has been a fun space. This is going to get, we're in Henryville, California. And when we were looking at that, at various space, you know, it's very hard to find the kind of space that has that is conducive to making. Hardware.
And a lot of, you know, we're not looking for a space that a law firm would occupy. We're also not necessarily looking for a space that is a true industrial space. You're looking for something that's kind of that hybrid where it's got, it has industrial aspects with respect to power and so on. But it's also something where people can reasonably work. I think we've hit the balance reasonably well. The, the heat is, is, has been an adventure to get this place properly heated. I can't imagine.
Yeah, and you know, it's, it's the Bay Area. So, you know, I know we talked a lot about HBO Silicon Valley a lot. And I'm sorry to go back to that well again. But there's so many just like nuances of HBO Silicon Valley that are just absolutely genius. And one of them is, there's an episode in which it's cold.
You have to understand in the Bay Area is like 57 degrees, which for most of the country would be like 57, I mean, 57 degrees is like, okay, it's like maybe got a little chill to it, but like there are plenty of days that I would love for it to be 57 when it's, you know, Steve went to school in Madison and there are plenty of days in Madison where you're like, I, I, I, I, I live in Toronto. So that is still a short cent tanked off whether.
Yeah, and in the Bay Area because we're so spoiled stupid, like we don't have double-pained windows. We don't have any of the percol, any of the things that you would normally need to survive and other climates. So, every one just like freezes their tails off when it's, and there's a, there's an episode of Silicon Valley where everyone is wearing like parkas inside because it's so cold. Right, exactly. And it's like actually, that's just not that cold.
But then, so this office is very like on brand for the Bay Area in that like it's, they're days in which we've had employees who have worn winter hats while working here. If we only knew someone at a heating air conditioning company. That's right. We can try this off for that. Well, thank you for giving us a tour of the office space. I think it's really cool. Like, first, so I studied electrical engineering and this really reminds me of a lab and that is just what really exciting.
You actually feel like a real engineer, not a fake engineer, to just wear the keyboard and a screen. So, it's, it's, it's pretty amazing what you've done with this space. And I think the rack part that you mentioned, so pretty much, if someone gets this rack, they pretty much need to, well, unbox a rack but then plug it in and it's ready to go. More or less. That's right. Yeah, that's pretty incredible.
Add power, add networking and go, which was the intent because I mean today it takes, it still takes companies on the order of two, three months from the time that all the boxes land at the warehouse to being able to get them unboxed. God forbid that you've got the wrong rack rails that shipped, even if you've nailed everything else in terms of getting server assembly storage assembly, networking assembly.
The facilities person that comes in because they can control being able to wire these things up and then getting all the software licensing taken care of. And you can have been invoiced for something and have an invoice do and still be months away from developers being productive, which is the exact opposite of the public cloud experience. Right. Right. Right. And go. And it a lot of it is prime mentioned, I mean, even just the crate.
And what we we had to go through to make sure that we had a repeatable process for creating and shipping to make that unboxing and that final assembly, which is, you know, if you want doors on the side and in side panels and doors and those sorts of things, all of that shipping with the rack itself, do we not want to tour the conference room? I mean, this is so exciting. So exciting. The one the one the one drawback of the space was not not a ton of separated off space, but it works well.
I think it's really good for said, Individi, like people will be going around and just I don't know why boarding together. This sounds fun to me. Yeah. And it's been a good space. Yeah, it has been a good space.
I mean, I think that the, the, it's interesting the image of the serenity because you know, you see all of this kind of the, that the return to office maintenance, which we all kind of rolled our eyes on because like the reason we need the physical aspect of the office is not for the collaboration for it is, it really is for the physicality of what we're building.
And the fact that we are people like you are a hardware company, I mean, we're hardware and software systems company, but that like, but you're also like all remote. It's like, well, yes, I mean, a lot, yes, there's a lot of lab work that's involved in hardware. There's also most of the work is not most of what our W's do. You are interacting with software.
You're actually with EDA software, simulation software, you're, you know, you're in solid works or you're in an ancestor or you're an Altium. And you, or you're, you're collaborating with your peers and that collaboration process on. So, or you were going to a manufacturing site to do bring up, you are going to like, there's, you know, it very, very hard to go build a company like this where everyone's in one spot.
And let's not forget, I mean, offices are valuable. You've got child avoidance that is. Her amenity that is actually critical. I mentioned last time we're expecting our first baby in three weeks. So I'll take that. Congratulations. Congratulations. I keep hearing that from people. You were going to come to the office after that. You will, I thank God it's Monday for sure. The yes. Yeah, that's great. And that's your full congratulations. Yeah, it's exciting.
And I think that it will be, you know, every eight would part of what makes parenting interesting is that just when you get really like you're just at your wits end, things change. So just remember that when you're at your wits end. Or when you get comfortable, just when you get comfortable, you're like, oh, I figured this out. Now we know how to keep him or her asleep. Yeah. And like, wow. This is so exciting.
I would say that there is a parenting is the is in the other ways to do this, but parenting is extraordinary for your own sense of empathy. And I think it changes like the way you think of other people when you've got it, you got this other person that you are going to love unequivocally despite all their faults and it's good. It's good for your everything. It's good stuff. For sure. So I want to question one. Go ahead. Go for it. I was going to talk about the crate. So I was going to say that.
That is such an attention, like a nice example of y'all's attention to detail because I think it's like, well, like 9,000 pounds. Y'all say something. It's a crate. I think. I know the showcase with the rack in there, but I think I can. Yeah, you should be able to pop it. Actually, it's a it's a. I should come out there if you want to pop it. Sure. Do we want to see the inside? Yeah, we did. If you should do a decreating experience. I think that would be an arizing video.
Yes. And we can make it. Yeah, that's actually. We should. Okay. So Steve is. Our lovely host. Yeah. Thank you, Steve. Going. Yes. So I'm very curious because Steve is going to do something that I would not have done, I guess. Or I guess it's that that guy. There's a lot of sort of. They're just hinges. This is the part of the. So. And. Um. Yep. Oh, wow. Oh, that looks pretty cool. All right. So that's that is the oxide rack as it arrives.
And then you can see that we've got a built-in ramp there. All right. You can just drag it out. Oh, yeah. So. Yeah. So this is where you get to like all these little bits that are engineered, right? So like the. So those. Those bolts come undone. That. That small piece comes out. And then the whole rack rolls out. Side panels are shipped are along with it. So. Nice. These are the these are the side panels here. Um, the you've got the the front and rear doors. Wow. And this particular.
Oh, God. Oh, wow. That looks heavy. Yeah, exactly. You can see. Yeah, so. You know, wheel it out. So that gets wheeled out and then. So that's a it's it's on wheels. And then once you actually get it in its destination, then then you you will. These will actually screw down. So you can start. Yeah. Yes, you can. Sorry. Terrible analogy probably. But looking at this, it reminds me of that. Using an oxide server rack is probably easier than assembling IKEA furniture. Yes. Uh, it is.
So no, it is definitely. It is easier than assembling IKEA furniture in that the it actually is. It's kind of funny to think about that way, but the in part because it all comes pre-assembled. Right. So it was very important that we were able to. To ship this thing completely intact to optimize for for for that time. Um, the. Yeah, yeah. So yeah, the idea here is of course that that rack will get gets wheeled into your data center.
On plug in those power whips and configure networking, which is not which shouldn't be trivialized. I mean, there's a lot to actually eat in and the whips too, right? The power is not is is not trivial. Yeah, this is very cool. Thank you for showing us what was inside of the of the great. So we're touch on create a little bit.
And so you last time we talked, um, you mentioned that you wrote, you know, your own networks, which this time you mentioned that, uh, you wrote your own manufacturing software. You're also designing the crate. Um, in building the sort of rack, you're also designing not just parts of the right, but everything around it too. And I want, I was curious to know, um, if there is one goal that you're optimizing for, like all of this is an optimization of something.
If I think of social media companies that are optimizing for the amount of time you spend on the screen, um, search, not necessarily any particular engine, but any search engines goal is to take you off of that pay just when it's possible. So in this specific case, what is that one goal that all of this is an optimization for?
Well, I think it's, I'll give my answer. Uh, I think what we talk about this end and process of manufacturing, doing our manufacturing software, which by the way requires having the right partner in manufacturing that is open and willing to you doing your manufacturing software.
And I don't just have like the agency and the ability to go kind of control the entire end and process if you don't have the right partners. And I think we talked a little bit about the Sanio, danky partnership last time we met and how important that was for us to be able to customize and just the fan speed as an example. But I mean, at the end of the day, what we're, what we're striving for is delivering the benefits of cloud computing on premises.
And so what does that mean? Well, a lot of what we just spoke about is time to value one of the one of the paramount things of public cloud computing is that 16 digits of freedom. It's like I swipe a credit card. I have access to this elastic infrastructure that allows me to go start deploying software that is this idea I have or this job I have to go complete.
And I don't have to think about any of the rest and it's fast. That is the exact opposite of that on premises experience. So one of the paramount goals is to collapse that time to start from 100 days to, you know, I mean seconds once this is installed.
That installation process to go from months down to hours. And to do that, you know, why do things take longer? Well, obviously we talked about the assembly, all the integration, the assembly and the checking and the testing that one has to do before they can turn this over to development teams.
And then another is quality. What is how do you get to know good quality of system before of the complete system before it's being installed because the worst place to find out that something is wrong is that the customer site, you know, just generally.
And so back to like kind of the manufacturing software, you know, being able to integrate that indoor end and process just gives us better visibility and confidence about the quality of system that is going to get plugged in and be able to test the end end system, not test individual components, the server separate from the switch separate from the control plane, but being able to kind of do that end and test.
And one, I mean, two of the very important kind of higher level goals are delivering that efficient experience to customers and then delivering that that kind of high level of quality of product. Certainly for things like the crate manufacturing software, when we talk about kind of some of the some of the things we took on ourselves.
And that's exactly it. Right. And so that's our lens is, you know, we're not seeking to actually do everything on our own. And we would be, you know, we want to take off the shelf solutions where we can. But there are a couple of these points that are the, they're actually the off the shelf solutions are either poor or they're expensive or they're so proprietary. We can't meaningfully integrate with them. I mean, so manufacturing software is a good example where the manufacturing software is.
It's kind of ridiculous that anyone would buy off the shelf manufacturing software because you manufacturing process is so, I mean, it is at some level idiosyncratic to what it's unique to what you're building. And you want to be able to control all the aspects of that. And, you know, I think that our contract manufacturer definitely thought we were, it's like, okay, you guys are going to, okay, you really do need to do everything on your own.
But when we, boy, talk about something that we've got absolutely no regrets over. That was infatically the right decision. That's enabled us to control every step of the process. And the thing that we're always braced for, it hasn't happened to date. But the thing that you want to get ahead of is that you do have a manufacturing quality issue that God forbid does show up in the customer base.
And it's really important that you were able to get on that very, very quickly. And you want to be able to, from that first failure, be able to realize, oh, we've got a bad lot. And we've got a, and now work with our partners to figure out how did we get here. And then, and then improve the process such that it doesn't happen again.
And that comes from us controlling this, the whole thing soup the nuts that also, by the way, allows us to do it because we don't have 18 different vendors in here slicing away at our margin. It allows us to deliver a product that is way more cost effective for our users. So, I mean, economically, it's very important that this, this has to deliver terrific value to our customers.
And it needs to do that, and it needs to do that in a way that we can take responsibility for any problem that that rack actually has. And that's the lens by which we make all these different decisions. And honestly, like a bunch of these decisions, we were, we didn't make them, you know, we came to it reluctantly, things like manufacturing software, or like doing the switch even.
I mean, we did not, when we first set out, we were not going to, you know, doing our own switch is really, is really complicated. But it, we also came to the conclusion that it was absolutely necessary. And in hindsight, is essential. It's not something we could have partnered with. I think you hit on a super important one, which is being able to take ownership of that entire customer experience.
And that, and then we talked about this last time. But when, just when you think about doing all the hardware and software together in product form, one of the biggest driving forces there was to make sure that if customers have an issue, we can take complete ownership of it. Whether it's part of our product or not. And to do that, you need to have, you know, be thoughtful about how you get telemetry throughout the system.
And also that you don't have parts of the product that are dependent upon third parties, where you don't need to. So that when there's that issue, you avoid the classic kind of gaslighting of like, I don't know, what, you know, what X are you running on this? Or what are you individual customer doing that is creating an issue that we don't want to take responsibility for.
It's quite the opposite being able to say, you know, to Brian's point of its manufacturing issue, if it is kind of anywhere in that end to end system, ensuring we can we can take ownership of that and correct that issue for the customers is maybe the biggest driving factor to areas where we hit that fork in the road of like, build versus buying. And we definitely look back with no regrets on the things that we have taken, taken the troll of in that process.
I think the aspect that you mentioned that you have reluctantly come to that conclusion. I think that word reluctantly is very important. Because it's very easy to get distracted by many of these things because from an engineering standpoint, it's like, oh, it'll be fun to solve this problem. And sure we can do it, Vesher.
But then if it's not in the service of your overall goal, it can end up distracting people a lot, which I mean, all of us have seen happening enough whether the team is big or small, whether it's a company young or big. And I think it comes down to having that balance in terms of where you are intentionally making that trade off. We'll spend more time here, but it's going to be worth it. But I'm glad to hear that these decisions have worked out so well for you.
Yeah, and I think you're exactly right. I think it is because it's overly reductive to say, oh, well, of course the engineers want to build their own crate. Like every engineer fantasizes about building their own crate. And it's like the so you can't assume that when you when you got a group of engineers saying, I think we actually need to write this ourselves.
I think it's a mistake to assume that like, well, of course, they're saying that because oftentimes engineers are coming to that out of the like we are trying to use this other thing. And we are and it's part of the reason why it's actually helpful to have a disposition of like, all right, we are going to try to use this other thing, but we're going to keep an open mind.
And if we begin to hit friction here, like we're going to want to like go reevaluate that determination a little bit. And you know, a lot of the things that we have gone our own way on are because we tried to not go our own way. And came to the just we got lucky enough that come to an early enough conclusion that like, no, actually we can't.
And then we're also were we are willing to take on the stuff ourselves and we are willing to be like, yes, we will do our own switch will do our own network operating system. We'll do our own embedded operating system like we've got the boldness to do that where it makes sense. We're not, but not for hearty.
So speaking of the server rack so back in June of last year. So this was like four years after founding outside you guys ship the very first server rack. This was a huge moment. And I think someone already was forced to either shoes. Just kidding. But not really. So then I was listening to a awesome friends episode that you guys recorded like right after and Brian. So I loved how you describe the moment of like a finally shipping being a bit like surprising.
So I'm paraphrasing here. But you said it was like it's like climbing a mountain and there's so many false summits and you get to a point where you're just grinding and all of a sudden you're at the actual summit and you're like, oh wow, like we're actually here. Like, were there any what were the sort of the memorable like false summits for you guys?
Well, because I think and when you are on a peak with many, many, many false summits at some point like you just break along the way and you're just like, okay, they're all false summits at this point. And I mean, that's kind of an important mental breaking to go to because you really just need to be focused on the I've got to go take the next step.
And the but no, there were so many times when the and it wasn't so much that like the boy we thought this was the end and it wasn't so it's not that this was the false summit metaphor breaks out a little bit. But you have so many experiences along the way that when you're building an artifact like this, nobody wants a machine that doesn't boot.
It's not like that's an and it's not like, oh, that product is like less desirable. It's like that product is got is though zero desire ability. Nobody wants that. And there are so many things along the way that like if we don't resolve this problem, it's not like we've got a product that's not quite as good.
We don't have a company like we are actually dead if we don't resolve this problem. And you know, you in those moments, you always need like you need some number of folks obviously that are headstown engage on the problem. Then it's also helpful to have some folks that are just missing just enough contacts that they can remain totally optimistic of like, oh, well, we'll nail this. And it's like, okay, glad you're optimistic.
And this is true not just on the engineering side, but certainly on the go to market side too that you know, like we've got the, you know, we've got that customer that was really excited that has now gone totally silent on and has not replied to the last like, you know, three emails.
And you know, this is where I'll be with the Steve, I'm like, like, like, they said like that they're really interested and they've got a project and seems like, oh, glad you're so optimistic. Like do you want me to explain to you? Like, do you want me to talk you out of your optimism? And I think that you need both, right? You got to be, you need to have that kind of paranoia and know all the ways that like, you kind of take it seriously.
Right? And like, you need to, though, like, this is an existential threat that we actually can't get this thing, can't get this Nick to come out of reset or, you know, can't, and that's kind of an obvious existential threat, but it's also as much of an existential threat.
And like, no, like, I know we were, like, this customer was very, or this prospect was very excited, but has now gone silent. And it's like, actually, there are lots of reasons to be paranoid about what's going on. And we need to make sure that like we're on that. When, you know, you've got a customer that's deployed, you've got to, you can't take that for granted, you know, I kind of at every stage.
And if you've got something that is like a whisp of smoke at on a production deployment, like, you need to be jumping all over that because, and then you kind of need that balance of, because it's very easy in any of those moments to be like, we're done. It's like, it's over. Everybody go home. This is, you know, I, what the hell are we thinking that we could pull this off? Like, no, it's like, fold up the tent.
And that's, there's a part of your inner voice that's doing that. And I think that for across the company, I think we are all, you know, domain experts and veterans. So we all have that own, our own inner voice in our own domain. But we can also be optimistic. I think it's like, we don't, we'll get through this. And I think as a company, we've not so many of those down now that I've got a much more, I would say robust sense of optimism.
And I'm less plagued by the idea that like, we'll never pull this off because we now overcome so much that we can actually go, we can do this stuff. And we want to not be, you can kind of become foolhardy in those moments too, where you get, you know, overly bold. So you have to like retain that fundamental humility. But fortunately, reality has a way of continuing the dish that out to us. So starting with a little bit of delusion and then working towards optimism.
And 100% and I mean, this is like the whenever you're doing something new, like you've got to balance the world as you're seeing it, the world as it could be with the world as it is actually. With like, no, but actually we like we don't. I know I can see great vision. Meanwhile, that it's like actually like, you know, this customer's got totally silent on us. Meanwhile, back in reality. So it's like you got it like any and it's it's at that intersection that you kind of have to constantly live.
So as we think about the entire development process and eventually shipping the product to a customer now having customers agreed that they want to buy something like this and actually pay money for it, not just say, oh yes, I'm interested. You require someone to talk with the customer. So this is where the go to market team comes into picture.
So we touched on this last time, but we didn't get to talk enough about it. So we wanted to get your take on two aspects here and you can go any other direction in whichever order you see fit. But the two aspects are when you're starting to build the company, you haven't necessarily validated the product itself meaning you don't have a working product yet.
So at what point do you start building a go to marketing and then when you decide you need to build it, how do you actually go about building it? Yeah, I mean, and I think we touched on it a bit last time, but this is possibly the number one place where companies fail.
Like if you if you kind of like walk backwards from a company that didn't make it very often, I would say, you describe, you know, a high percentage of it is from thinking that you're further along than you are and starting to pour a bunch of money in.
And you know, I think there's plenty of blame to be left at the feet of VCs in this regard because where a lot of that emphasis and pressure comes from with a lot of young companies is from the investors that have been, you know, and their minds patiently waiting for a long time for this thing to be ready that they poured a bunch of money into and took a bunch of risk in and had to make the case internally into their LPs and
everything else, but you know, there's this anticipation of, okay, it's it's go time. And you know, I'm speaking now about many colleagues and many past companies where they have gotten to a point where yes, they're now shipping a product, but they know that they still need to get kind of affirmation and also just prove points back in terms of what aspects of the product.
And then quickly optimizing that product to eventually find patterns and build those patterns out and everyone kind of lazily collects this into product market fit, but it's a whole, you know, set of stages that incorporate finding repeatability in the sales process, the deployment process, the retention of customer and expansion of customer process.
But back to that moment where there's the launch and there's the pomp and circumstance and, you know, the PR firm is out beating the streets and it's like, let's go, let's time to gas it. And that is where I think a lot of companies, understandably, are feel that they need to. They need to to kind of live up to what everyone expects of them and that can be an extraordinarily expensive mistake.
Because you are just shoveling cash into the furnace with the expectation that you're going to be able to start scaling some of the early signs that you have seen. And, you know, now back to oxide, in our particular case, we had an even earlier stage problem, which is we had a lot of potential customer interest and we were still years away from having a product we can sell to someone.
So in our case, it was, you know, at what point in the development process do you start investing in people that are going to begin developing these customer relationships and you have to, you do have to start developing those relationships long before you're selling a product because in particular with the markets that we're selling to, you know, the kind of the fortune 1000 space and regulated industries. You don't have the privilege of just showing up and saying, okay, we're ready.
You know, write a big check. You know, you have to have really understood how that company operates. And, you know, you have to have gone through the fundamentals process of discovery and understanding, you know, what are the things that are driving their business? And what are the kind of the highest order strategic comparatives in the company and then how do those map to the product that you are delivering and or what sorts of pain points are you solving for?
So we did invest in that early, but kind of the approach that we took was trying to find someone who had the kind of had the versatility to come in and be a pioneer. Someone who was sort of unafraid of blazing some tough trailheads and that was able to be equally comfortable in a conversation where they were kind of the business sales front door for oxide, but also had the depth and credibility of a technologist.
And so we were kind of looking for what some companies will call like a sales engineer, solutions engineer, but who was, you know, interested in experience in engaging with larger enterprises. But then we had kind of the next part of the equation that we had to figure out, which is like also find someone that was going to be willing to do this work for a year plus without having anything to sell. Which is sounds kind of adds.
Yeah, adds a little bit of complexity to the search because there are great and I think this is where, forgive me, get on my sales soap box, but I think the sales industry generally can catch some flags slash be like routinely denigrated for some of the bad patterns that exist in sales and there are plenty of them and that is very well warranted in many, many, many cases.
But there also is and I've why it has been so deeply enjoyable to work, be able to work close to the brine of the last 15 plus years is seeing actually how many surprising patterns there are between sales and engineering.
And it is when when done right sales is a real craft and it's a craft that is born out of for the when done the right way it starts with deep, deep, deep curiosity about everything around someone and really deep curiosity about how products work and you know how technology works and then deep curiosity about how this particular company ticks.
And wanting to understand what are the different constraints and pressures that are going on in this company and then in this person's role and and and what are the things that you really want to understand if you are going to qualify mutually qualify whether this particular product is a good fit for them and vice versa.
And and so it is a but but at the same time then you have the other side of sales which is much more transactional and and what sales done wrong is more predatory and it's it is it unfortunately can also stem out of curiosity but it's a curiosity around how you can convince someone to do something.
Which which is kind of the perverted sort of the dark side of where curiosity can go wrong and and and there are amazing amazing sales people that are very good at that they're very very good at that and they are very good for their management team and their investors and and so and and then you've got everything in the middle right which which is you have good sales people that are good for certain stages of companies that are good when there is.
Product market fit and they're very very good at scaling and organization at implementing a system that allows the provides a foundation for being able to bring in more team members and do it in a repeatable manner that allows you to start really building that customer engagement across a much bigger market but winding it all the way back we needed this pipe here we needed this person that was willing to to to deal with that.
And so we had to deal with that kind of nuanced period before we really had before we had a product in market and long before we had product market fit and and yet this person was going to be super super low bearing because they were collecting all of this important data that was going to inform kind of in some cases you know decent shifts in kind of how we were getting to finished product and in other cases just slight little tweaks on what we were building.
Up and down the stack and but we but we needed to make sure that we did not over emphasize that we couldn't afford to go invest in three or five or ten of these folks until we were getting closer to launch and I have to say I think we found a pretty darn good one.
We found someone who had really good technical depth and experience in the networking world had been at Cisco for a number of years, a risk of for a number of years had kind of worked with companies big and small and it was it was a higher that we had to get right and it really helped the engineering team the product team kind of shape what we were doing. And how we're getting that product to market. And Travis who's that yeah I can even call him by name exactly.
But Travis also I think it's got a really it's got exactly the right disposition which is a tough balance to head of like you need to be optimistic and forward looking but you can't have happy years and happy years is and there are plenty of
because sales folks can be kind of naturally optimistic and they get this kind of natural enthusiasm that can be infectious when they are actually not reading the customer exactly right and they think there's an opportunity that's actually that's bigger than it actually is or they think that they have not read what they haven't totally understood what the customer is trying to do completely.
And they earnestly not out of malice but they just they they end up believing that things are further along than they are and in some cases where they're very charismatic they can end up closing a sale that makes no sense. And that's like well okay but the sales of sale it's like well not not really and because if you if ultimately what you're leaving in your wake is a customer that's dissatisfied or has something that actually don't need.
They like that's a problem that's not a way to actually build a sustainable business like yes you got to sell. Congratulations you got a quarterly number up there but you haven't actually because you because you're not on the right foundation and you know as an extreme example of that you know at at joint we took our software and that we were running our own cloud on and we sold it to other folks so they could run that they could run a last infrastructure on their computers.
And we signed a new customer it was very very excited to be a joint customer we were very excited and I was it's like okay we closed the sale great and I'm on kind of like the first call to figure out what this actually looks like.
And I'm like okay so I mean one thing that's important is like you know what are the actual machines that you're running on that's like I don't know like whatever machines you all ship me and like the computers that you buy we make the software you buy the hardware and they're like wait what?
Like so they literally had been led to believe and and this is one of these things where it's like I don't think that they necessarily been lied to I think that they had they had an enormous misunderstanding they were not corrected. They they were not corrected. And this is like you know this is where it's like you're in a gray area right where it's like you know is that is it dishonest to sell to someone who has a grievous misunderstanding of the product.
It's like well dishonest is probably a bit strong but it's like not prudent to do that and you're going to end up in this case you're like the customer's like well okay wait a minute then like I'm paying that amount just for the software.
This is like a lot more expensive than I thought it was because I mean you're just like oh my god and you can be a course you're just like how are you possible how have you managed to like safely get out of bed in the morning if you got such because like the customer's grievous misunderstanding at some level like this is on that right.
But the ultimately it was not a sale and because ultimately it's like that's not what they wanted and it was a huge waste of time and energy and effort and it came out of big time happy years. But it's a by it's a byproduct of a bigger problem though and the bigger problem was that that was that was one example of what was a rush into go to market.
It was a company that raised a lot of money and in raising that money were kind of influenced by those investors some of which had a stake in the joint turning from a public cloud computing company to an enterprise software company.
That was going to then go kind of quote unquote be the arms dealer to telcos everywhere that wanted to build their own AWS's and that is an interesting idea and financially it sounds really good because now you can kind of be selling the picks and the shovels to this growing global world of federated clouds and yet it had not been tested.
So before testing going and hiring a giant go to market sales army and marketing engine because you can't sell without marketing and I say that with tongue squarely planted cheek and you which by the way like marketing is its whole whole other topic which is incredibly important and when done right is company making but
you know building a top down sales organization the moment that you've got a good idea and you've got a little bit of market kind of like indicators and tailwind and themes is the kind of classic pattern that ends up not working very often.
And you know you can do any of these to the extreme that that but I think it is it is super important and really healthy to rather than that pattern of like okay and this is a pattern that we saw a lot of last 10 years like two technical co founders they look at each other and be like oh well who's going to be the CEO I don't know I don't want to talk to you you talk to me you you you want to deal with those like investors and all those like I want to I want to go build stuff and that was like okay yes.
Sure I'll do it and then start building the engineering team and start building the product and getting some good signal from customers and then I you know it's sometimes getting gas a little bit but like getting told that like hey you need to bring in that that sales leader.
Like you got to have the business person needs to be brought in and that is a point in which I think you know those those founding teams need to think very carefully about who that next hire is and you know I will always make the case that that next hire should not be the VP of sales.
That next hire should be that on the ground pioneer who is going to go out and learn as much as possible work learning from the engineering team and the product team learning from these perspective customers that have come in enthusiastic and are like oh we totally need your product and you know some of the best selling is is digging in and understanding that that customer doesn't need your product and getting to a short no instead of a long no.
And and that's best for both parties and that you are going to hire a better VP of sales with that next three to six months of intelligence that you can go gather and the the concern around the opportunity cost of like yeah but if we don't do it now these things take so much time you know even to find a VP of sales to hire an on board and then for them to build their team because they don't do the I mean they're not in the car they're not in you know in the meeting.
Like for them to build the team that is going to go get then on board and warmed up and you know we're not going to see real repeatable revenue for for 12 months 18 months if we even get started today we can't afford to wait and it's like it's actually the that is the moment that your company hangs in the balance and what you need to do is if you want to think about it this way one way to convey this to to your investors is like what I heard it's going to be a good thing.
What what I heard at just a couple months ago with a bunch of folks that are running go to market and big you know what you would consider kind of like big big big startups now and they call the third VP of sales problem and the third VP of sales problem is that you is is not even that there were three in this particular company but in a company you've got the first VP sales and they do a bunch of like heavy lift hard work but guess what there wasn't repeatability there weren't real patterns or
real patterns or real kind of repeatable pain points and customer segments that they could go attack so they would get one Z2 Z sales all over the place and moving too slow and they got fired next VP of sales gets brought in and they kind of carry this momentum forward and they they get to a point where they're right on the precipice of success but again too slow they get fired and that third VP of sales comes in and just crushes now that third VP of sales
crushing the bigger problem that is building is for that next company that hires that person away as their CRO they found the absolute great white shark like we we found our we found our person and they hire that person based on a pattern of success that they had almost nothing to do with building and concrete example I we the we were and I'm going to all anonymize this just enough we were referred to someone who's like this person is extraordinary they built company X they they were
responsible for like they were bill company X from nothing and are responsible for companies at company X's success I'm like they arrived at company X after it was public I would compete it this is like this is you know years ago but I like I I know this company pretty well and it's like they so you looked at what they were claiming it's like they this this
company had a hundred million dollars to revenue when you showed up and it's like you didn't build that hundred million dollars revenue because you weren't there when that was built and there's a big difference between having a hundred million dollars revenue and having ten million
dollars revenue is big difference to having ten million dollars revenue having one million dollars revenue and the kind of the but the person that was referring to this was like believes in their heart that that individual was responsible for the success of that company
even though the company they arrived at it already enjoyed quite a bit of success that they were not responsible for it all and so this happens a lot that third VP of sales that third VP of sales and that's the one who doesn't necessarily know that they've kind of forced forced into this thing and it's like it's like they also may believe in their heart that they themselves are responsible for it all and you know I always the opening of the
Simpson's you know where they've got Maggie's got the little steering wheel in the car that march is driving and I always think about that little steering wheel so frequently and what's going on in Maggie's car yeah Maggie's like I'm driving and it's like no no you we've really provided you the delusion that you're driving and I feel that so often that third VP of sales I'm not to take
anything away from them who can get who and it certainly may have had the prudence to realize this is a company that was that was in the right position but it's like very often it's like there's a bunch that came before you that is responsible for laying the foundation that you are now believing that you created well and I mean and again then to go back it I think it is not a question of should this early stage
technology company higher VP of sales it is just making sure that they are thoughtful about the timing and the steps that should in most cases this is of course nothing is uniform for all companies but it really should make sure to to learn as much as they can to hire the right one it's just it's really really important to find the
characteristics that are going to fit for the stage of the company and so you also don't get a situation where you did hire a good one but they bounce out in 18 months because you are prepared for them and so but it's it's it's hard I mean of course it's hard you've got all this pressure companies that have raised money there are expectations where once the product is done you've got to go scale
and and it is one of the trickier periods in a company's life cycle because if you get it wrong it is actually like an 18 month pause on the business because you have to go rip all that up again and then go like rebuild that again when you have the right fit and but but when you have the right fit it's gold because now to speak to the value of a good sales leader in in in an organization you know the these folks have a depth of experience
around building the fundamentals for repeatability in the sales process in customer engagement building the right culture so that you have a culture that is anchored around cost like responsibility for customer and doesn't think about it in a transactional manner where it's like to Brian's point like oh I hit my quarter I threw my deals over the fence now I'm on to like the next you know the next ones that I want to go land and hit my number again
that that are thinking about how these customer relationships are not measured in 90 days or 180 days better measured in like five years 10 years and when that is the case when you have a culture where folks are thoughtful about the reputation and relationship as you can imagine there's more empathy for the customers there's more curiosity and discovery and and and there's more discipline around when having tough conversations where it isn't a good fit for a company
a good fit for a particular customer and helping them understand that you know a perspective customer and so you know there are there are really important there's really important foundation that a good sales leader puts in place for a company
it it's just that so often we just see so many so many so many examples of companies either rushing in or being rushed in to it to that and then being entirely differential to that to that all knowing person and it takes a very very long time and a lot of money to realize how costly that was so I I both can can tell the cautionary tale and also I think again when done right it can be a company making higher it's a it's a very very important part of the part of the team
as you were describing some of the traits of good sales leader or even the first person that you hired to get a lot of that data in the first six months for example I was just thinking about in my head for engineers or for companies who only have technical co-founders initially and don't have the sales person on the team as you mentioned CEO whoever maybe they flipped a coin and said this person is going to the CEO
they can also learn from the characteristics you described to play that role for the time being and a couple things that I gathered from what you said is one is just curiosity I think we touched on curiosity as a trade lost time to like just be that curious
yes it's so it is so imperative and it is it's it's kind of it's hard to overstate how important it is and I guess I've been frustrated recently so I think that when you look at a company that has a great technology but is struggling to actually build a business around it
you often have behind that company folks that are actually in curious about how how it's being used and so we saw a whole and sorry this is this is kind of under my fingernails at the moment because we there was a whole era where you had these open source projects that got that tried to build companies around those projects and the projects themselves were wildly popular but in the exceptions to this but if you look you can go through there many of these there's not just one of these
or many of these examples where if you talked to the people who are behind the open source project they weren't that interested in why this thing was popular how it was being used or how they would actually in their mind like the project is the product and it's the project is like super interesting to folks that's why people were using it but the product they actually wanted they wanted something more than that for the product
and the you know that again there have been a bunch of these and I think that if you are and people have wondered it's like well okay so we can't actually ever turn it's impossible to turn an open source project into a to a product
and it's like well it does require like more work and it especially requires some real curiosity about like what does someone want around this often like by the way like some the things that people want are like hey yeah I would love to have like LDAP integration or this
80 integration or I actually it's very important to me that the quality of the support experience is terrific around this and if you were the you know you're one of these VC funded open source projects you're like oh like quality of support like really that's what I mean like a D integration like I don't know that just does not sound that's not that exciting well okay then you should not just like don't go into business like that's like that's fine just don't do just the
or find someone who is more interested in because you've created something that people are using get curious about why they're using it and how it is a part of the thing that they're building get interested in that because it is interesting I think this is you know that I think that that curiosity is something that really binds us as a company like I mean like if you're the kind of person who are like if you're going to like nod off at a factory tour not be interested in you know
someone taking you like I'm going to take you inside of my business and show you what my business looks like that should be really interesting to you and especially if that person be like oh and by the way here's how your thing I envision your thing might fit into the thing that I'm doing that should be like lighting you up and if it doesn't you should really question whether you should be starting a company around this.
So many engineers who are not necessarily finding a company right now starting a company right now but they're working at another company and in their heads that like at some point in the future I definitely want to start one and I think at that point for them to be in a position where they can do a little bit of the go to market work and they don't have let's say the right product here or they don't have the salesperson yet.
What are some of the things that they can start doing from even now at their current job to just kind of groom the skill a little bit because this takes a lot of practice and I think all engineers would be better off if they were good at talking and it's selling I would say and selling is a skill that I think everyone should have but what are some things that engineers can do to get better at this?
Yeah well I think last time Steve suggested taking someone to lunch and understanding what their pain points are or someone else in the company. I think that remains really good advice. I think the other advice that I would add to that is go take a customer to lunch. Go sit with an actual customer. You know if you're an established company, great news.
That company has product market fit at some level like that's putting I mean I don't know maybe you're at one of these like VC fuel disasters that like literally has got no customers but hopefully not. Hopefully you've actually got customers and like go sit with those customers and understand how they're using your thing and why and you know even if that's just confirmatory even if it's just like oh this is why the thing that I'm making is important to you as an end user of this thing.
Make those connections and I think in some companies that I think these companies do themselves an enormous disservice. I think Google does this infamously where it's really really hard for a Google engineer to go understand what a Google user is going through.
But it's like come on you're Google you're smart go figure it out like go find one like you know go like yes you're going to I get it then you're going to go you know fight through some brush to get there but go get there go to a customer and get again get curious about how and I think that's a way that you can start to build that. You're also going to see just a Steve said last time you're going to see a whole bunch of problems that you yourself may have perspective on.
And I think I was going to give a similar answer on go go meet with customers and when you meet with customers you know at minimum figure out you know how they are using your product and why they chose the product in the first place and what they considered. In terms of alternatives what were the alternatives why did they not choose the alternatives.
And that will start to elucidate like some of the thought process of what goes on behind the walls when a company is contemplating making a buying decision on anything and then even better if you can get into understanding what that person's world looks like personally and not personally personally but like personally in that workplace.
And what are the things that make them doing their job difficult that maybe specific to that company but oftentimes these are pretty repeatable patterns where it's like yeah well you know unfortunately the finance team just doesn't understand the value of you know what this brings to the business. And it's like well how do you how do you articulate that value who articulates that value.
And like well that's the problem is honestly like no one can they're just literally looking at the bottom now I'm denigrating finance there's a lot of finance incredibly important as well but but but then you know or you're getting an answer where it's like oh well then this person in the organization is who has to go deliver that message.
And that is incredibly important information to be gathering about how these bigger organizations work and getting into some of the fundamentals about you know what is the buying process in an organization and who are champions who are coaches who are the decision makers who are the economic buyers. And when you kind of come back to building a good sales culture in in that company that this person has now started and they're looking at how do we go start to scale this.
There are some real frameworks that you want to to get in place so again you have repeatability and and but back to that person at the company today there's so much opportunity to learn from from your current customers that's where. And so the definitely is the best time spent for sure and also just understanding things that have nothing to do with the product like you know you go read the 10k of that customer and just look at the risk section that get called out in the earnings.
Like because again these are the things that are weighing on the executive team that are getting pushed down in the organization and sometimes that person that you're having lunch with actually doesn't understand why this person is requiring a certain team to do things a certain way.
And there are plenty of cases where like there's perverse incentives or perverse punishment that's coming down from pissed off investors or a board or these kinds of things that you can actually kind of glean out of those quarterly earnings reports that there's all public information if you're curious.
I try to do like a version of this not great at my first job and then sort of the push back I think from like management was sort of like yeah do not put an engineer in front of the this was like enterprise sats and say do not put an engineer in front of the customer and then you know let him go on about you know stuff. So that was a bit hard I think to just like directly trying to because you know these are like farm or like kind of like I feel like.
So a suggestion there if you're if one finds themselves in that same situation there's an easy answer which is like don't put me in front of a customer put me in front of a perspective customer that you think we have 0% chance of closing. Put me in front of someone that you think is like a dead lead I mean again there's a certain if the person's non responsive it's kind of hard to you know but hey who is someone we talk to that went with somebody else.
And I will tell you an engineer has a much higher probability of getting a next conversation with that person than anyone in the sales organization and you're going to be able to and you've got this credible like hey I'm an engineer I built the product reaching out because I would just I'd be so deeply grateful if I could get any feedback about how you think about just the problem space.
We know that we're not going to be a fit for you all like you're not going to be a customer but but man it would be helpful to understand this and you may have to reach out to four or five of these dead prospects to get one to agree to take a meeting but oh man you walk back with the information learned in that meeting and the CEO like anyone of the organizations like where is he we need him in the room right now so we can figure out like what we did wrong in this case.
And you're going to get the other side of it you're going to get why the customer didn't buy and it may be product related it may have nothing to do with product it may be like honestly the rep didn't get back to me for seven days and by the time they did I was already you know down the path with this other person and so I would that's an easy one because the sales organization thinks that thing is not even a one percent chance it's a zero percent chance so why not what's the harm nothing good looks nothing like that.
So to your point and then so what I tried to do was then to basically go to the account management side and then meet friends with like a account manager and I sort of use that as like a proxy and that was yeah like to your point like super helpful I was like oh I had no idea like this is how like they use the tool I was like very clueless but yeah like learn the tongue from just like having even that perspective as a proxy.
So quick time check we have about eight minutes I had three questions to turn down but I think we can only do one to raffle so I'll I'll list the questions and you get to pick. Okay so one question I had was around transparency and sharing challenges with we see is in team members so for example when you're building the product at times as you said the machine is in booring so you don't have a product in which means you don't have a company.
How transparently do you share that with the team and the we see this well because I've seen two versions of this not necessarily in terms of the startup but even big companies this glass of optimism which everyone can see through and like that's bullshit or the other version is being very transparent and sharing that good.
That's a strong we can knock that one like 60 seconds I think yeah so I mean so you've got to love the bad news yes and you have to you know you know good I've got good news and bad news everyone should want the bad news first.
Like give me the bad news and we you need to be in an organization which it which includes like your potential investors and board where I know I want the bad news I want to know the bad news first like the and you want to be able to like you know give me enough context so I you don't want to just like induce panic and everybody all the time but it's important that we have the bad news and that you're able to get so you need to be transparent about that.
You can you can be transparent but but bring the bad news with one of two things either a concrete ask of who you're bringing the bad news to of some action you want them to take or with a having already thought through mitigation options for the bad news so either come with like here's the path we're going to be on you do not need to take action or I need you to take action and here's the action I need you to take the worst possible.
The worst possible thing you can do is just drop that news on the table and just say like I don't know we got all the good news. I know I like to tell me you love bad news. It's a little more nuanced. I love that. I love that. Action of the bad news. Action of the bad news are having the mitigation for it. I think that makes perfect sense.
And I think that in terms of like our investors for example there's been plenty of times like look here's the issue we've got here's what we've done to mitigate it here's what works like this is purely informational you don't and I mean I think virtually those conversations are always like okay thanks for letting me know that this is obviously very high priority. You know keep me abreast of any details and you know right that's what we want.
Good advice. Okay next to let's see if we can do those two. The first was GPUs so right now I think the oxides over that comes with CPUs only AMD CPUs are fired. And you see so much investment going on just building out data center capacity with GPUs is that direction that you're planning to pursue with oxide.
It's definitely interesting I mean obviously and the accelerated compute is extremely important and there are different ways in which it's important these training workloads are less a little less interesting to us for a bunch of reasons in part because the real value that we bring at oxide
holistic approach ease of management multi tendency running the stuff on promises it that is a little these training workloads don't look like that training workloads are kind of these dragsters that you take out with ultra high speed networking huge amounts of GPUs and data the GPUs like are popping all the time you've got reliability problems everywhere they are then doing a bunch of things to kind of compensate for the reliability problems.
And then like when you it's like Bitcoin mining like you're done with it you can just like push this infrastructure into the ocean I don't care by the way I need twice as much money to build the next one. And like that is a little less those workloads are a little less interesting and also like the folks that are that are really running those workloads aren't oxide customers anyway there they're their own hyperscage right there.
And they're the ones are actually you know building their own power plants and so on that said I think what's very interesting to us is on the inference side. And I think there's something that we're very interested in is the what does it look like when we're doing a lot more inference in a lot more parts of the stack.
And that looks like having I think we're going to see the middle get muddy with respect to CPUs versus GPUs CPUs adopting instructors at extensions that allow that inference to happen on CPU. And then also going to APUs where we have these you got cores on die that are accelerated cores living alongside general purpose course is very interesting to us.
So this is the the MI 300A from AMD is a super interesting part to us we love to run is that is the next generation AMD part that has things like AVX 512 that was an instruction set extensions to kind of put some concrete meat on the bone there and that's really I think that you know if you if you assume. Which I think is reasonable to assume that the rise of LMS generative AI there's a lot that's real there right this is not crypto money there's the there's a bunch that's real there.
Because it's real we are in this is going to be a multi decade trend this is not something that is going to evaporate in the next two years so we are much more interested in what does the intersection look like when this is in that kind of mainstream on prime enterprise. And to us that that's got to look less like these kind of ultra specialized workloads and it needs to look more mainstream with respect.
And it needs to look less like repackaging of GPUs in one you chassis and has to look more like rack level infrastructure that includes CPU and GPU and the networking substrate and rack level power management and a control plane that goes from workload to all the resources under. And you know the Nvidia's definitely demonstrated this playbook with the acquisitions that they have made that they want to own all five of those points and be able to deliver a rack level solution around it.
And the good news such as it is is that we've been at work for five years now on four of those five a cloud computing control plane rack level power management and networking management controlling the network.
Having the CPU by way of obviously our partnership with AMD and we just now need to figure out what the right hardware acceleration component is to to integrate in to deliver a robust solution that is going to fit as Brian mentioned this more general purpose AI landscape in the next couple of years. So we'll definitely will definitely be participants for sure it's just a question of doing it the right way.
And unfortunately that question was too meaty so we're now out of time for the third question we get we get another six minutes. We have six minutes. Okay. All right. The next question kind of ties into this one which is I saw the street recently where someone mentioned that the see this friend where companies now want to move out of public cloud where like if you look at the last 10 years everyone wanted to get into public cloud.
But I think base camp is one of the examples where this started blogging about their like this is how much it costs we are going to move out and run our own cloud and it costs a little. So curious to get your perspective like do you see this trend overall in the customer she talking to as well.
Yeah. I think the trend that is absolutely the case is that five years ago everyone got to the end of their probably five year evaluation of public cloud and said public cloud computing as a architecture is the future. Like this notion of having programmable API driven elastic infrastructure resources I want that I want that foundation for all of my software deployment and management and operation.
And so because of that I am now going to tell my board I'm going to tell my investors that we are shutting down 80% of our data centers and we're moving all this to the public club. And and and what's happened over the last five years is in various states these companies have gotten 20% of the way through that project and been like whoa wait a minute. Why does this not match the spreadsheet that I presented three years ago cost or data transfer is surprising one to many people.
Yeah. What is this light item that I don't remember seeing this on the proposal. But the economics of renting are are difficult for vast large persistent workloads over you know longer periods of time. And that does not take anything away from the value of cloud computing. I think that is not what gets mistaken in this narrative right now of the cloud repatriation of like oh moving off the public cloud we don't want the public cloud anymore has absolutely nothing to do with the product.
And has everything to do with the difficulty trying to push the entirety of one's business into that model into a rental model into a model where it runs in you know eight regions instead of 80 locations. And so where a model of share everything not share everything but a multi tenant model that regulators are not totally comfortable with yet or that creates a a vector for attack that people are not comfortable with whether that's reasonable or not.
So I think reasons that include economics governance not being able to control and understand what the dollars are being spent on security latency is is why these are slowing. And it's actually a slowing in migration to the public cloud more than it is everyone running for the exits. And people saying wait a minute I'm I'm sold on the benefits of cloud computing is an architecture.
I can't put as much there as I had hoped so now what do I do what how do I how do I square this and then yes at the same time you also have a group of kind of a part of the industry that maybe grew up 100% in the public cloud and are looking at the economics or the lack of control over the quality of the product underneath them.
The inability when their customers are pissed off because something went bump in the night to tell them like why it did when it will happen again what they're doing to mitigate it you know unfortunately when you when you are outsourcing all those layers.
You're not going to have the best answers to your customers when they're frustrated about something so maybe it's wanting to take more control over the experience delivered to your customer maybe it's because you want to you want to lower latency maybe it is economics there are another group of companies that are saying hey being 100% public cloud born is not enough.
And all the way to the extreme of like I'm leaving the public cloud but I think you know 98% of this driving force of on prem kind of coming back into favor and and this conversation happening about wait a minute not only is it not dead it's like this could be one of the biggest growth areas ever in the next 10 years is is much more around people trying to answer the question of how do I get the benefits of cloud computing everywhere my business needs to run.
And how do I if I'm a cloud sass born platform how do I double the size of my business in the next 10 years it's going to be tough to do that if my total addressable market is only the data people put into the public cloud. So how do I go extend my platform to reach all the data on the planet and to do that you you have to have this hybrid model where you're using public cloud using someone else's cloud infrastructure on a rental basis when an economically makes sense.
And you're using your own be you know I mean the reason we start this company is that you should be able to own the computers that run your cloud for the rest of the use cases where from a perspective of economics or security or latency or other makes the most sense and when you're doing that you shouldn't have to have this this fractured front end to the system you should have this same elastic infrastructure resource and API driven kind of primitive.
That is in both in all sides rented or we're owned that's very well put I know we can keep going but we're at a time Steven Ryan thank you for being so gracious with your time we love our conversations with you and hopefully again we'll try to get in the future at some point but again thank you so much for today and also thank you for showing us the office space plus the rack that is super cool.
Okay we have to we have to end where we started last episode in the the meet cute which we can't begin with a person who had reached out to Dell with an inquiry and had submitted a web form to see if they could get someone's attention at Dell because some micr systems was not paying them any attention or sufficient attention.
And here's this going. We're at the all just wait what wild coincidence or stroke of fate while we were recording this that same person reached out to oxide and and that person's reach out to oxides starts with the following I had met Steve about 20 years ago via a web form I was hoping to do the same again we're looking for a VM or replacement and we're going to be
scaling our control plane across more sites so we are we're again going to be reaching out and and it it it was a nice nice close loop here. Yes. On the first second podcast episode incredible incredible thanks for sharing that. Yeah thanks for that. I'm going to get you get your podcasts to any other company updates that you want to share with me via a podcast. Oh now that you mention it.
Yeah. Hey thank you so much for listening to the show you can subscribe wherever you get your podcasts and learn more about us at software misadventures.com. Until next time take care.