Cloud Commitments Without the Lock-In with Archera's Aran Khanna

I was at I was at Costco and I bought I I was at I was at Costco and I bought I want to say 900 want to say 900 want to say 900 of Advil for like 25 bucks. of Advil for like 25 bucks. of Advil for like 25 bucks. And then I was at the Hilton a couple of And then I was at the Hilton a couple of And then I was at the Hilton a couple of weeks ago and I bought two Advil weeks ago and I bought two Advil weeks ago and I bought two Advil for $5. It was unbelievable. So, I'm for $5. It was unbelievable. So, I'm for $5. It was unbelievable. So, I'm starting to think that like on-demand starting to think that like on-demand starting to think that like on-demand pricing is the convenience store price pricing is the convenience store price pricing is the convenience store price of compute of compute of compute and not the real price. and not the real price. and not the real price. >> I think that's that's generally it, >> I think that's that's generally it, >> I think that's that's generally it, right? It It's almost the promotional right? It It's almost the promotional right? It It's almost the promotional price. That's That's where you want to price. That's That's where you want to price. That's That's where you want to prototype at. That's where you want to prototype at. That's where you want to prototype at. That's where you want to like understand, hey, you know, this like understand, hey, you know, this like understand, hey, you know, this works. But then if you're actually going works. But then if you're actually going works. But then if you're actually going to run the thing, to run the thing, to run the thing, >> Hi, I'm Scott Hanselman and this is >> Hi, I'm Scott Hanselman and this is >> Hi, I'm Scott Hanselman and this is another episode of Hanselminutes. Today, another episode of Hanselminutes. Today, another episode of Hanselminutes. Today, I'm chatting with Ajeet Khanna. He's the I'm chatting with Ajeet Khanna. He's the I'm chatting with Ajeet Khanna. He's the co-founder and CEO of Arrcus. How are co-founder and CEO of Arrcus. How are co-founder and CEO of Arrcus. How are you, sir? you, sir? you, sir? >> I'm doing well. Thank you so much for >> I'm doing well. Thank you so much for >> I'm doing well. Thank you so much for having me. having me. having me. >> So, I as a individual developer who does >> So, I as a individual developer who does >> So, I as a individual developer who does like side projects and startup like side projects and startup like side projects and startup consulting, I kind of pay as you go. And consulting, I kind of pay as you go. And consulting, I kind of pay as you go. And I've been doing pay as you go in the I've been doing pay as you go in the I've been doing pay as you go in the cloud for gosh, cloud for gosh, cloud for gosh, 17, 18 years. But I'm realizing that if 17, 18 years. But I'm realizing that if 17, 18 years. But I'm realizing that if I were to get serious about the cloud, I I were to get serious about the cloud, I I were to get serious about the cloud, I could probably save a ton of money if I could probably save a ton of money if I could probably save a ton of money if I would just commit.

would just commit. would just commit. But then I'm realizing that cloud But then I'm realizing that cloud But then I'm realizing that cloud commitments are like this financial commitments are like this financial commitments are like this financial instrument that's hiding inside of a instrument that's hiding inside of a instrument that's hiding inside of a developer infrastructure and I just developer infrastructure and I just developer infrastructure and I just don't want to don't want to don't want to buy a year's worth of compute. buy a year's worth of compute. buy a year's worth of compute. And I Arrcus seems to have a solution And I Arrcus seems to have a solution And I Arrcus seems to have a solution for that for for you know, for at least for that for for you know, for at least for that for for you know, for at least most businesses. Is that right? most businesses. Is that right? most businesses. Is that right? >> Yeah, absolutely. As you said, you know, >> Yeah, absolutely. As you said, you know, >> Yeah, absolutely. As you said, you know, commitments commitments commitments are almost on their face to some extent are almost on their face to some extent are almost on their face to some extent antithetical to what the cloud promise antithetical to what the cloud promise antithetical to what the cloud promise was to developers like you and me, which was to developers like you and me, which was to developers like you and me, which is, you know, pay for what you need, is, you know, pay for what you need, is, you know, pay for what you need, turn it on, turn it off, and you know, turn it on, turn it off, and you know, turn it on, turn it off, and you know, pay as you go. But pay as you go. But pay as you go. But you know, my background is I actually you know, my background is I actually you know, my background is I actually spent almost a decade working for the spent almost a decade working for the spent almost a decade working for the hyperscalers, working for Azure, working hyperscalers, working for Azure, working hyperscalers, working for Azure, working for AWS, and launching a lot of these for AWS, and launching a lot of these for AWS, and launching a lot of these services. And what you realize is services. And what you realize is services. And what you realize is there's a ton of capital expense that there's a ton of capital expense that there's a ton of capital expense that goes into these projects and so having goes into these projects and so having goes into these projects and so having committed revenue from customers is committed revenue from customers is committed revenue from customers is actually incredibly valuable and in fact actually incredibly valuable and in fact actually incredibly valuable and in fact as the cost of data center builds go go as the cost of data center builds go go as the cost of data center builds go go up, it's actually more and more up, it's actually more and more up, it's actually more and more necessary. And the main way that necessary. And the main way that necessary. And the main way that historically cloud providers have historically cloud providers have historically cloud providers have incentivized customers who were really incentivized customers who were really incentivized customers who were really sold on this pay as you go model as you sold on this pay as you go model as you sold on this pay as you go model as you and I were when we first started in the and I were when we first started in the and I were when we first started in the cloud building and experimenting and cloud building and experimenting and cloud building and experimenting and doing hobbies was to give deep discounts doing hobbies was to give deep discounts doing hobbies was to give deep discounts for actually committing to for actually committing to for actually committing to infrastructure versus running it on infrastructure versus running it on infrastructure versus running it on demand. You know, these things can go demand. You know, these things can go demand. You know, these things can go north of 80% if you take something from north of 80% if you take something from north of 80% if you take something from a, you know, second-by-second usage a, you know, second-by-second usage a, you know, second-by-second usage pattern to a 3-year committed usage pattern to a 3-year committed usage pattern to a 3-year committed usage pattern. And most developers, obviously, pattern. And most developers, obviously, pattern. And most developers, obviously, especially if you're running a small especially if you're running a small especially if you're running a small hobby, you don't know what's going to hobby, you don't know what's going to hobby, you don't know what's going to happen with it, um they don't feel

happen with it, um they don't feel happen with it, um they don't feel really comfortable committing. They're really comfortable committing. They're really comfortable committing. They're they're willing to pay the premium. But they're willing to pay the premium. But they're willing to pay the premium. But when you get to running a business on when you get to running a business on when you get to running a business on the cloud, especially a serious business the cloud, especially a serious business the cloud, especially a serious business running at scale, you know, we work with running at scale, you know, we work with running at scale, you know, we work with some of the biggest cloud customers in some of the biggest cloud customers in some of the biggest cloud customers in the world here at Archera, uh you have the world here at Archera, uh you have the world here at Archera, uh you have portfolios of hundreds of millions if portfolios of hundreds of millions if portfolios of hundreds of millions if not even now billions of dollars not even now billions of dollars not even now billions of dollars committed to the cloud providers. Um and committed to the cloud providers. Um and committed to the cloud providers. Um and it's become, you know, one of the the it's become, you know, one of the the it's become, you know, one of the the largest problems that we see in the largest problems that we see in the largest problems that we see in the cloud ecosystem today, not just the the cloud ecosystem today, not just the the cloud ecosystem today, not just the the technology problem, but the financial technology problem, but the financial technology problem, but the financial problem that precipitates from it. problem that precipitates from it. problem that precipitates from it. >> Yeah, it almost seems like we as >> Yeah, it almost seems like we as >> Yeah, it almost seems like we as developers were so enamored with the developers were so enamored with the developers were so enamored with the idea of this infinitely scalable elastic idea of this infinitely scalable elastic idea of this infinitely scalable elastic cloud that we kind of forgot it's still cloud that we kind of forgot it's still cloud that we kind of forgot it's still hosting and there's still machines there hosting and there's still machines there hosting and there's still machines there and if the machines aren't spinning, and if the machines aren't spinning, and if the machines aren't spinning, then who's paying for all of that and then who's paying for all of that and then who's paying for all of that and they're being wasted. So then that means they're being wasted. So then that means they're being wasted. So then that means that if I'm doing pay as you go, I'm that if I'm doing pay as you go, I'm that if I'm doing pay as you go, I'm effectively paying a huge premium for effectively paying a huge premium for effectively paying a huge premium for the elasticity, for that bursty kind of the elasticity, for that bursty kind of the elasticity, for that bursty kind of thing. I I And as a startup, you would thing. I I And as a startup, you would thing. I I And as a startup, you would think I would want to actually pay less.

think I would want to actually pay less. think I would want to actually pay less. Like you said, 80%. I mean, I've seen Like you said, 80%. I mean, I've seen Like you said, 80%. I mean, I've seen 50, 60, 70% like crazy upfronts. 50, 60, 70% like crazy upfronts. 50, 60, 70% like crazy upfronts. Buying a VM is pretty reasonable, but Buying a VM is pretty reasonable, but Buying a VM is pretty reasonable, but buying it for a year is like you pay for buying it for a year is like you pay for buying it for a year is like you pay for 4 months and the rest is free. I'm 4 months and the rest is free. I'm 4 months and the rest is free. I'm leaving money on the table, but then I I leaving money on the table, but then I I leaving money on the table, but then I I worry that what if my company goes under worry that what if my company goes under worry that what if my company goes under or what if what if I need more or what or what if what if I need more or what or what if what if I need more or what if I need less? I don't I just don't if I need less? I don't I just don't if I need less? I don't I just don't want to spend I want to spend the least want to spend I want to spend the least want to spend I want to spend the least amount possible. amount possible. amount possible. >> That that's exactly it. And what's >> That that's exactly it. And what's >> That that's exactly it. And what's really interesting, you know, you talked really interesting, you know, you talked really interesting, you know, you talked about that illusion of infinite capacity about that illusion of infinite capacity about that illusion of infinite capacity before that existed in the cloud. Now before that existed in the cloud. Now before that existed in the cloud. Now with with tokens and GPUs and capacity with with tokens and GPUs and capacity with with tokens and GPUs and capacity shortages in AI like that illusion shortages in AI like that illusion shortages in AI like that illusion itself is almost broken, right? We see itself is almost broken, right? We see itself is almost broken, right? We see that often in 2026 you're required to that often in 2026 you're required to that often in 2026 you're required to get a commitment just to have access to get a commitment just to have access to get a commitment just to have access to something like a latest generation or something like a latest generation or something like a latest generation or even an older generation GPU nowadays even an older generation GPU nowadays even an older generation GPU nowadays given the the capacity crunch. And and given the the capacity crunch. And and given the the capacity crunch. And and that's really what Archera is kind of that's really what Archera is kind of that's really what Archera is kind of holistically helping customers solve holistically helping customers solve holistically helping customers solve for. If they need to commit to get for. If they need to commit to get for. If they need to commit to get access to capacity or if they're access to capacity or if they're access to capacity or if they're interested in getting the discount, but interested in getting the discount, but interested in getting the discount, but they have uncertainty. They don't know they have uncertainty. They don't know they have uncertainty. They don't know if they're going to be using this GPU if they're going to be using this GPU if they're going to be using this GPU for the whole year. They don't know if for the whole year. They don't know if for the whole year. They don't know if they need all this compute. You know, 3 they need all this compute. You know, 3 they need all this compute. You know, 3 years on you know, their business if years on you know, their business if years on you know, their business if they're a startup might not even be they're a startup might not even be they're a startup might not even be around or it might be doing something around or it might be doing something around or it might be doing something completely different. And what we do is completely different. And what we do is completely different. And what we do is we help them hedge that downside risk we help them hedge that downside risk we help them hedge that downside risk by a unique product. We're the only ones by a unique product. We're the only ones by a unique product. We're the only ones in the space offering it called insured in the space offering it called insured in the space offering it called insured cloud commitments, which is essentially cloud commitments, which is essentially cloud commitments, which is essentially as simple as saying, "Hey, if you commit as simple as saying, "Hey, if you commit as simple as saying, "Hey, if you commit for 3 years, but you don't use say the for 3 years, but you don't use say the for 3 years, but you don't use say the last 24 months, we'll pay you back the

last 24 months, we'll pay you back the last 24 months, we'll pay you back the difference for that whole 24-month difference for that whole 24-month difference for that whole 24-month period if you're basically having a period if you're basically having a period if you're basically having a commitment sitting there in your commitment sitting there in your commitment sitting there in your inventory that's not being used. And inventory that's not being used. And inventory that's not being used. And this gives customers the confidence to this gives customers the confidence to this gives customers the confidence to commit with that downside risk protected commit with that downside risk protected commit with that downside risk protected and capture more of that discount, you and capture more of that discount, you and capture more of that discount, you know, get resources that actually know, get resources that actually know, get resources that actually require commitments they otherwise require commitments they otherwise require commitments they otherwise wouldn't be willing to commit to like wouldn't be willing to commit to like wouldn't be willing to commit to like GPUs. Um, and really provide a new GPUs. Um, and really provide a new GPUs. Um, and really provide a new option they can put into that overall option they can put into that overall option they can put into that overall commitment portfolio. What we see is commitment portfolio. What we see is commitment portfolio. What we see is that customers generally, you know, the that customers generally, you know, the that customers generally, you know, the most sophisticated biggest ones that we most sophisticated biggest ones that we most sophisticated biggest ones that we work with, um, you know, like Hex has work with, um, you know, like Hex has work with, um, you know, like Hex has become a really big AI native customer become a really big AI native customer become a really big AI native customer that have been running with us for 7 that have been running with us for 7 that have been running with us for 7 years. You know, traditionally it would years. You know, traditionally it would years. You know, traditionally it would be a mix of on-demand, 1-year, and be a mix of on-demand, 1-year, and be a mix of on-demand, 1-year, and 3-year commitments. You know, now when 3-year commitments. You know, now when 3-year commitments. You know, now when they're mature, they're actually looking they're mature, they're actually looking they're mature, they're actually looking at a mix of on-demand, insured 1-year at a mix of on-demand, insured 1-year at a mix of on-demand, insured 1-year commitments, regular 1-year commitments, commitments, regular 1-year commitments, commitments, regular 1-year commitments, insured 3-year commitments, regular insured 3-year commitments, regular insured 3-year commitments, regular 3-year commitments. It becomes a lot 3-year commitments. It becomes a lot 3-year commitments. It becomes a lot more of a dynamic portfolio that helps more of a dynamic portfolio that helps more of a dynamic portfolio that helps them really match the risk in their them really match the risk in their them really match the risk in their business with the kind of best pricing business with the kind of best pricing business with the kind of best pricing that they can get out of these cloud that they can get out of these cloud that they can get out of these cloud provider constructs.

provider constructs. provider constructs. >> So, the cloud is an abstraction layer, >> So, the cloud is an abstraction layer, >> So, the cloud is an abstraction layer, and as a developer, I get to pretend and as a developer, I get to pretend and as a developer, I get to pretend that there isn't capital expenditure. I that there isn't capital expenditure. I that there isn't capital expenditure. I don't get to pretend that the giant don't get to pretend that the giant don't get to pretend that the giant building in the giant data center building in the giant data center building in the giant data center doesn't exist. I just, you know, I I doesn't exist. I just, you know, I I doesn't exist. I just, you know, I I burst when I want to burst. But, let's burst when I want to burst. But, let's burst when I want to burst. But, let's pretend for a second that I've never pretend for a second that I've never pretend for a second that I've never heard the words FinOps or reserved heard the words FinOps or reserved heard the words FinOps or reserved instance. Maybe instance. Maybe instance. Maybe talk to me like I'm five. talk to me like I'm five. talk to me like I'm five. There's a giant Costco with a bunch of There's a giant Costco with a bunch of There's a giant Costco with a bunch of computers in it, and they're not all computers in it, and they're not all computers in it, and they're not all working very hard. working very hard. working very hard. Why is it okay for them to give me 80% Why is it okay for them to give me 80% Why is it okay for them to give me 80% off, and why do they want reserved off, and why do they want reserved off, and why do they want reserved instances? instances? instances? >> Yeah, so, you know, a core part of this >> Yeah, so, you know, a core part of this >> Yeah, so, you know, a core part of this just goes back to just goes back to just goes back to how these services, like you said, get how these services, like you said, get how these services, like you said, get built. You are actually going and buying built. You are actually going and buying built. You are actually going and buying machines and putting them into a data machines and putting them into a data machines and putting them into a data center, which takes a lot of money up center, which takes a lot of money up center, which takes a lot of money up front. In fact, the best analogy I found front. In fact, the best analogy I found front. In fact, the best analogy I found in an explain-it-like-I- I'm five sort in an explain-it-like-I- I'm five sort in an explain-it-like-I- I'm five sort of way is looking at this almost like of way is looking at this almost like of way is looking at this almost like building an apartment building. You building an apartment building. You building an apartment building. You know, we all live somewhere. You know, know, we all live somewhere. You know, know, we all live somewhere. You know, there's a big apartment building down there's a big apartment building down there's a big apartment building down the block. It takes a lot of money to go the block. It takes a lot of money to go the block. It takes a lot of money to go and build these apartments, but then you and build these apartments, but then you and build these apartments, but then you got to sell them. And you could put them got to sell them. And you could put them got to sell them. And you could put them up on Airbnb, you'll probably get the up on Airbnb, you'll probably get the up on Airbnb, you'll probably get the best price, but you might not fill it, best price, but you might not fill it, best price, but you might not fill it, you know, the whole year, right? You're you know, the whole year, right? You're you know, the whole year, right? You're going to have a lot of vacant apartments going to have a lot of vacant apartments going to have a lot of vacant apartments if you're only Airbnbing. And you might if you're only Airbnbing. And you might if you're only Airbnbing. And you might not recoup that cost. And in fact, you not recoup that cost. And in fact, you not recoup that cost. And in fact, you generally have a loan against that generally have a loan against that generally have a loan against that apartment, and they don't want to, you apartment, and they don't want to, you apartment, and they don't want to, you know, underwrite Airbnb revenue. They know, underwrite Airbnb revenue. They know, underwrite Airbnb revenue. They want to underwrite, you know, 1-year, want to underwrite, you know, 1-year, want to underwrite, you know, 1-year, 3-year apartment leases, where there's a 3-year apartment leases, where there's a 3-year apartment leases, where there's a guarantee of spend every single month guarantee of spend every single month guarantee of spend every single month coming in, not some variable expense.

coming in, not some variable expense. coming in, not some variable expense. The same way these cloud providers go The same way these cloud providers go The same way these cloud providers go and take loans, you know, CoreWeave is a and take loans, you know, CoreWeave is a and take loans, you know, CoreWeave is a very extreme example, but all of them do very extreme example, but all of them do very extreme example, but all of them do this, this, this, to go and put up GPU capacity, to build to go and put up GPU capacity, to build to go and put up GPU capacity, to build these data centers. They don't want to these data centers. They don't want to these data centers. They don't want to just rent it out on demand with no just rent it out on demand with no just rent it out on demand with no guarantee. In fact, they probably can't guarantee. In fact, they probably can't guarantee. In fact, they probably can't secure their loans if they do that. They secure their loans if they do that. They secure their loans if they do that. They want to have those 1-year and 3-year want to have those 1-year and 3-year want to have those 1-year and 3-year leases. And the same way as when you leases. And the same way as when you leases. And the same way as when you rent an apartment and you sign a 1-year rent an apartment and you sign a 1-year rent an apartment and you sign a 1-year lease versus a 3-year lease, you're lease versus a 3-year lease, you're lease versus a 3-year lease, you're going to get a better rate going to get a better rate going to get a better rate if you sign a 1-year lease on on a if you sign a 1-year lease on on a if you sign a 1-year lease on on a database, you're going to get a better database, you're going to get a better database, you're going to get a better rate than running it on demand. But if rate than running it on demand. But if rate than running it on demand. But if you sign a 3-year you sign a 3-year you sign a 3-year lease, you're going to get the best lease, you're going to get the best lease, you're going to get the best rate. And then really the way to think rate. And then really the way to think rate. And then really the way to think about Archera and all of this is about Archera and all of this is about Archera and all of this is thinking about essentially having the thinking about essentially having the thinking about essentially having the right to break that lease. So let's say, right to break that lease. So let's say, right to break that lease. So let's say, you know, get a new job in Memphis, you you know, get a new job in Memphis, you you know, get a new job in Memphis, you know, you want to get an apartment know, you want to get an apartment know, you want to get an apartment there, but you need to sign a lease. You there, but you need to sign a lease. You there, but you need to sign a lease. You don't know if you're going to keep the don't know if you're going to keep the don't know if you're going to keep the job after a year, but you sign a 3-year job after a year, but you sign a 3-year job after a year, but you sign a 3-year lease. You got a great rate. We'll lease. You got a great rate. We'll lease. You got a great rate. We'll essentially sell you the right to break essentially sell you the right to break essentially sell you the right to break that lease after that first year if your that lease after that first year if your that lease after that first year if your job circumstances change. Means you can job circumstances change. Means you can job circumstances change. Means you can get the the right capacity, you know, get the the right capacity, you know, get the the right capacity, you know, that right apartment. You can get it at that right apartment. You can get it at that right apartment. You can get it at a much better price than say Airbnb the a much better price than say Airbnb the a much better price than say Airbnb the whole time that you're in Memphis, whole time that you're in Memphis, whole time that you're in Memphis, but you have that downside protection but you have that downside protection but you have that downside protection baked in.

baked in. baked in. >> So is this a trick or is this do the >> So is this a trick or is this do the >> So is this a trick or is this do the cloud people know you exist? Because if cloud people know you exist? Because if cloud people know you exist? Because if I break my lease, the landlord would be I break my lease, the landlord would be I break my lease, the landlord would be upset with me, but if the landlord knew upset with me, but if the landlord knew upset with me, but if the landlord knew that I had Archera insurance, they might that I had Archera insurance, they might that I had Archera insurance, they might be okay with it. be okay with it. be okay with it. >> Yeah, and and that's exactly how we've >> Yeah, and and that's exactly how we've >> Yeah, and and that's exactly how we've gone to market, actually. We've gone to market, actually. We've gone to market, actually. We've partnered incredibly deeply with all of partnered incredibly deeply with all of partnered incredibly deeply with all of the hyperscalers, AWS, Azure, Google. In the hyperscalers, AWS, Azure, Google. In the hyperscalers, AWS, Azure, Google. In fact, you know, most of our revenue fact, you know, most of our revenue fact, you know, most of our revenue actually comes through selling on the actually comes through selling on the actually comes through selling on the cloud providers' marketplaces. cloud providers' marketplaces. cloud providers' marketplaces. So the Google, Azure, AWS marketplace So the Google, Azure, AWS marketplace So the Google, Azure, AWS marketplace is where we do actually most of our is where we do actually most of our is where we do actually most of our revenue and most of our volume. So we're revenue and most of our volume. So we're revenue and most of our volume. So we're really tightly integrated with them. And really tightly integrated with them. And really tightly integrated with them. And they love it. In fact, their sales teams they love it. In fact, their sales teams they love it. In fact, their sales teams love it because they are incentivized to love it because they are incentivized to love it because they are incentivized to sign bigger deals, to sign these sign bigger deals, to sign these sign bigger deals, to sign these commitments. And often it's very commitments. And often it's very commitments. And often it's very difficult to get customers to commit difficult to get customers to commit difficult to get customers to commit larger amounts and to commit for longer larger amounts and to commit for longer larger amounts and to commit for longer because of that downside risk aversion because of that downside risk aversion because of that downside risk aversion that they have. So, having something that they have. So, having something that they have. So, having something like this they could bring into a deal like this they could bring into a deal like this they could bring into a deal really helps in terms of closing deals, really helps in terms of closing deals, really helps in terms of closing deals, getting more velocity in terms of their getting more velocity in terms of their getting more velocity in terms of their contracts and what they're able to contracts and what they're able to contracts and what they're able to deliver to customers. Um and then deliver to customers. Um and then deliver to customers. Um and then obviously the customers love it because obviously the customers love it because obviously the customers love it because instead of sitting on something that's instead of sitting on something that's instead of sitting on something that's unused and getting mad at Amazon or mad unused and getting mad at Amazon or mad unused and getting mad at Amazon or mad at Azure, they actually have an out.

at Azure, they actually have an out. at Azure, they actually have an out. They can take those dollars and reinvest They can take those dollars and reinvest They can take those dollars and reinvest them in new initiatives. So, it's really them in new initiatives. So, it's really them in new initiatives. So, it's really value-added for the whole ecosystem. The value-added for the whole ecosystem. The value-added for the whole ecosystem. The same way that, you know, Ford probably same way that, you know, Ford probably same way that, you know, Ford probably wants Geico to exist, right? They want wants Geico to exist, right? They want wants Geico to exist, right? They want to have uh someone who insures these car to have uh someone who insures these car to have uh someone who insures these car purchases so people will actually purchases so people will actually purchases so people will actually purchase the big expensive F-150 in the purchase the big expensive F-150 in the purchase the big expensive F-150 in the first place. first place. first place. >> Now, I had a fender bender recently or >> Now, I had a fender bender recently or >> Now, I had a fender bender recently or my son did and there was an interesting my son did and there was an interesting my son did and there was an interesting discussion as I was teaching my son how discussion as I was teaching my son how discussion as I was teaching my son how insurance works and explaining the insurance works and explaining the insurance works and explaining the different flavors of insurance one could different flavors of insurance one could different flavors of insurance one could do. And he was wondering if the other do. And he was wondering if the other do. And he was wondering if the other person could have bought insurance and person could have bought insurance and person could have bought insurance and then pretended to backdate it and you then pretended to backdate it and you then pretended to backdate it and you know, basically game the system. So, we know, basically game the system. So, we know, basically game the system. So, we had conversations about insurance. How had conversations about insurance. How had conversations about insurance. How do you prevent abuse from someone who do you prevent abuse from someone who do you prevent abuse from someone who might know that their startup's going might know that their startup's going might know that their startup's going out of business and try to go and buy out of business and try to go and buy out of business and try to go and buy insurance and then and you know, make insurance and then and you know, make insurance and then and you know, make you insolvent? you insolvent? you insolvent? >> Yeah, absolutely. Well, it comes down to >> Yeah, absolutely. Well, it comes down to >> Yeah, absolutely. Well, it comes down to the core of insurance, which is the core of insurance, which is the core of insurance, which is underwriting, right? And part of underwriting, right? And part of underwriting, right? And part of underwriting is to avoid this thing underwriting is to avoid this thing underwriting is to avoid this thing that's called moral hazard, which is that's called moral hazard, which is that's called moral hazard, which is essentially where someone goes and buys essentially where someone goes and buys essentially where someone goes and buys insurance and then takes a lot of risk insurance and then takes a lot of risk insurance and then takes a lot of risk because they have insurance. So, there's because they have insurance. So, there's because they have insurance. So, there's a few kind of ways that we do this. I a few kind of ways that we do this. I a few kind of ways that we do this. I think the simplest way is essentially think the simplest way is essentially think the simplest way is essentially uh we ensure that the customer has some uh we ensure that the customer has some uh we ensure that the customer has some skin in the game. They get, you know, skin in the game. They get, you know, skin in the game. They get, you know, usage data when we work with a customer usage data when we work with a customer usage data when we work with a customer going back at least a year if not more going back at least a year if not more going back at least a year if not more and we can see what they've actually and we can see what they've actually and we can see what they've actually been spending money on, you know, in been spending money on, you know, in been spending money on, you know, in their cloud provider. We can see some of their cloud provider. We can see some of their cloud provider. We can see some of that history. And those are real dollars that history. And those are real dollars that history. And those are real dollars spent. So, they, you know, if they're spent. So, they, you know, if they're spent. So, they, you know, if they're going to try and trick us, they would going to try and trick us, they would going to try and trick us, they would have to spend a lot of money uh to do have to spend a lot of money uh to do have to spend a lot of money uh to do it. So, I think there's kind of a a it. So, I think there's kind of a a it. So, I think there's kind of a a default guardrail that we have around,

default guardrail that we have around, default guardrail that we have around, you know, when we even make an offer. you know, when we even make an offer. you know, when we even make an offer. Beyond that, we actually have, you know, Beyond that, we actually have, you know, Beyond that, we actually have, you know, now a petabyte plus of historical usage now a petabyte plus of historical usage now a petabyte plus of historical usage data across thousands of customers. So, data across thousands of customers. So, data across thousands of customers. So, we're able to actually go in and very we're able to actually go in and very we're able to actually go in and very granularly understand, "Hey, other granularly understand, "Hey, other granularly understand, "Hey, other customers that look like this customer, customers that look like this customer, customers that look like this customer, how long do they keep their, you know, how long do they keep their, you know, how long do they keep their, you know, RDS database up?" And we have very RDS database up?" And we have very RDS database up?" And we have very strong priors, is what we call them in strong priors, is what we call them in strong priors, is what we call them in the science world, on like what the the science world, on like what the the science world, on like what the longevity of these instances are. And longevity of these instances are. And longevity of these instances are. And then finally, there's kind of the last then finally, there's kind of the last then finally, there's kind of the last layer, which is almost in my co- founder layer, which is almost in my co- founder layer, which is almost in my co- founder came from the finance world. He was a came from the finance world. He was a came from the finance world. He was a D.E. Shaw doing, you know, quantitative D.E. Shaw doing, you know, quantitative D.E. Shaw doing, you know, quantitative pricing and and Uber after that. It's pricing and and Uber after that. It's pricing and and Uber after that. It's really this portfolio construction really this portfolio construction really this portfolio construction problem, because you actually want to problem, because you actually want to problem, because you actually want to build a very diversified book, where build a very diversified book, where build a very diversified book, where you're not overexposed to any one you're not overexposed to any one you're not overexposed to any one instance type, any one GPU type, any one instance type, any one GPU type, any one instance type, any one GPU type, any one customer type, or any one customer in customer type, or any one customer in customer type, or any one customer in general. Um and so, that kind of last general. Um and so, that kind of last general. Um and so, that kind of last layer, uh you know, those are kind of layer, uh you know, those are kind of layer, uh you know, those are kind of the categories of the guardrails that we the categories of the guardrails that we the categories of the guardrails that we have in place to ensure that we're not have in place to ensure that we're not have in place to ensure that we're not getting selected against, we're finding, getting selected against, we're finding, getting selected against, we're finding, you know, deals that make sense for both you know, deals that make sense for both you know, deals that make sense for both parties, and we're able to surface that parties, and we're able to surface that parties, and we're able to surface that up in the right way to end customers.

up in the right way to end customers. up in the right way to end customers. >> Is this that term I've heard quant, >> Is this that term I've heard quant, >> Is this that term I've heard quant, where there's a quant who is the person where there's a quant who is the person where there's a quant who is the person who is really doing the deep statistical who is really doing the deep statistical who is really doing the deep statistical analysis, and the more diverse your analysis, and the more diverse your analysis, and the more diverse your portfolio as an insurance company, the portfolio as an insurance company, the portfolio as an insurance company, the better? better? better? >> That's exactly it, right? We're we're >> That's exactly it, right? We're we're >> That's exactly it, right? We're we're taking a very data-driven quantitative taking a very data-driven quantitative taking a very data-driven quantitative or quant approach, and in fact, we do or quant approach, and in fact, we do or quant approach, and in fact, we do have folks who act as quants within our have folks who act as quants within our have folks who act as quants within our business, uh in terms of maintaining our business, uh in terms of maintaining our business, uh in terms of maintaining our underwriting algorithm, understanding underwriting algorithm, understanding underwriting algorithm, understanding where we should offer insurance versus where we should offer insurance versus where we should offer insurance versus pull it back, and and even thinking pull it back, and and even thinking pull it back, and and even thinking about like pricing very granularly. about like pricing very granularly. about like pricing very granularly. Where should we price a little bit more Where should we price a little bit more Where should we price a little bit more or price a little bit less, depending on or price a little bit less, depending on or price a little bit less, depending on the actual implied risk that we're the actual implied risk that we're the actual implied risk that we're taking. So, it's a very much part of the taking. So, it's a very much part of the taking. So, it's a very much part of the moat of this business, the fact that moat of this business, the fact that moat of this business, the fact that we've been doing this for 7 years and we've been doing this for 7 years and we've been doing this for 7 years and have such a rich data set, such a broad have such a rich data set, such a broad have such a rich data set, such a broad array of customers, and have really array of customers, and have really array of customers, and have really built expertise in this very niche thing built expertise in this very niche thing built expertise in this very niche thing that we're kind of the pioneer in. that we're kind of the pioneer in. that we're kind of the pioneer in. >> So, this might be a little spicy, but >> So, this might be a little spicy, but >> So, this might be a little spicy, but I'm sure you wouldn't appreciate a spicy I'm sure you wouldn't appreciate a spicy I'm sure you wouldn't appreciate a spicy question. Like, when did cloud billing question. Like, when did cloud billing question. Like, when did cloud billing first feel broken to you? And you know, first feel broken to you? And you know, first feel broken to you? And you know, was the pitch was only pay for what you was the pitch was only pay for what you was the pitch was only pay for what you use and I'm curious if you think at use and I'm curious if you think at use and I'm curious if you think at enterprise scale that's still true enterprise scale that's still true enterprise scale that's still true or this is the the correct solution to or this is the the correct solution to or this is the the correct solution to that broken problem.

that broken problem. that broken problem. >> Yeah, the cloud billing started to feel >> Yeah, the cloud billing started to feel >> Yeah, the cloud billing started to feel broken frankly when those two initial broken frankly when those two initial broken frankly when those two initial promises that we opened the show with promises that we opened the show with promises that we opened the show with you know the illusion of unlimited you know the illusion of unlimited you know the illusion of unlimited capacity and then you know the pay as capacity and then you know the pay as capacity and then you know the pay as you go nature started to break down. Now you go nature started to break down. Now you go nature started to break down. Now the original idea of like a reserved the original idea of like a reserved the original idea of like a reserved instance instance instance was interesting but it didn't feel was interesting but it didn't feel was interesting but it didn't feel unusual when it came out in the early unusual when it came out in the early unusual when it came out in the early 2010s from Amazon. You know hey this box 2010s from Amazon. You know hey this box 2010s from Amazon. You know hey this box is one that we're going to keep up for a is one that we're going to keep up for a is one that we're going to keep up for a while you know we should be consuming while you know we should be consuming while you know we should be consuming this in a slightly different way. But this in a slightly different way. But this in a slightly different way. But you know fast forward to 2017 when we you know fast forward to 2017 when we you know fast forward to 2017 when we were launching Sagemaker you know I was were launching Sagemaker you know I was were launching Sagemaker you know I was part of that original launch team over part of that original launch team over part of that original launch team over at AWS and we were saying look for at AWS and we were saying look for at AWS and we were saying look for specific skews we're not going to buy specific skews we're not going to buy specific skews we're not going to buy that much Nvidia gear. This is this is that much Nvidia gear. This is this is that much Nvidia gear. This is this is back then now I guess they're buying as back then now I guess they're buying as back then now I guess they're buying as much as they can and so we're going to much as they can and so we're going to much as they can and so we're going to have a limited on capacity so people are have a limited on capacity so people are have a limited on capacity so people are going to have to make a commitment just going to have to make a commitment just going to have to make a commitment just to show their bought in so we can give to show their bought in so we can give to show their bought in so we can give them some of that capacity. And so this them some of that capacity. And so this them some of that capacity. And so this idea of a commitment turned from idea of a commitment turned from idea of a commitment turned from something that could provide financial something that could provide financial something that could provide financial leverage and match sort of the customers leverage and match sort of the customers leverage and match sort of the customers consumption pattern you know a reserved consumption pattern you know a reserved consumption pattern you know a reserved instance to something that almost felt instance to something that almost felt instance to something that almost felt like a hammer that everything started to like a hammer that everything started to like a hammer that everything started to look like a nail with oh we have a look like a nail with oh we have a look like a nail with oh we have a limited capacity. Okay throw throw a limited capacity. Okay throw throw a limited capacity. Okay throw throw a commitment in front of the problem and commitment in front of the problem and commitment in front of the problem and that will fix it. Oh we want to actually that will fix it. Oh we want to actually that will fix it. Oh we want to actually stop customers from going multi-cloud stop customers from going multi-cloud stop customers from going multi-cloud let's make a big commitment code program let's make a big commitment code program let's make a big commitment code program like a private pricing agreement program like a private pricing agreement program like a private pricing agreement program and and throw a Microsoft Azure and and throw a Microsoft Azure and and throw a Microsoft Azure commitment program and throw everything commitment program and throw everything commitment program and throw everything into that. Oh we we you know we have into that. Oh we we you know we have into that. Oh we we you know we have issues around issues around issues around customers wanting to modernize their customers wanting to modernize their customers wanting to modernize their infrastructure let's just make a new infrastructure let's just make a new infrastructure let's just make a new commitment type call the savings plan commitment type call the savings plan commitment type call the savings plan and throw everything in that so you know and throw everything in that so you know and throw everything in that so you know this initial idea that started to make this initial idea that started to make this initial idea that started to make sense around a cloud commitment you know

sense around a cloud commitment you know sense around a cloud commitment you know more and more stuff got piled onto it to more and more stuff got piled onto it to more and more stuff got piled onto it to solve more and more problems and I think solve more and more problems and I think solve more and more problems and I think it started to become quite a spaghetti it started to become quite a spaghetti it started to become quite a spaghetti mess for customers to both understand mess for customers to both understand mess for customers to both understand and then ultimately to take advantage of and then ultimately to take advantage of and then ultimately to take advantage of in a way that made the most sense for in a way that made the most sense for in a way that made the most sense for their business. their business. their business. >> I was at I was at Costco and I bought I >> I was at I was at Costco and I bought I >> I was at I was at Costco and I bought I want to say 900 want to say 900 want to say 900 uh Advil for like 25 bucks. uh Advil for like 25 bucks. uh Advil for like 25 bucks. And then I was at the Hilton a couple of And then I was at the Hilton a couple of And then I was at the Hilton a couple of weeks ago and I bought two Advil for $5. weeks ago and I bought two Advil for $5. weeks ago and I bought two Advil for $5. It was unbelievable. So, I'm starting to It was unbelievable. So, I'm starting to It was unbelievable. So, I'm starting to think that like on-demand pricing is the think that like on-demand pricing is the think that like on-demand pricing is the convenience store price of compute convenience store price of compute convenience store price of compute and not the real price. and not the real price. and not the real price. >> I think that's that's generally it, >> I think that's that's generally it, >> I think that's that's generally it, right? It's It's almost the promotional right? It's It's almost the promotional right? It's It's almost the promotional price. That's That's where you want to price. That's That's where you want to price. That's That's where you want to prototype at. That's where you want to prototype at. That's where you want to prototype at. That's where you want to like understand, "Hey, you know, this like understand, "Hey, you know, this like understand, "Hey, you know, this works." But then if you're actually works." But then if you're actually works." But then if you're actually going to run the thing, you know, it's going to run the thing, you know, it's going to run the thing, you know, it's the imperative to start looking at the imperative to start looking at the imperative to start looking at commitments. And And maybe a one-year commitments. And And maybe a one-year commitments. And And maybe a one-year commitment doesn't make sense. Maybe you commitment doesn't make sense. Maybe you commitment doesn't make sense. Maybe you want to start with committing for want to start with committing for want to start with committing for uh a shorter amount of time and that's uh a shorter amount of time and that's uh a shorter amount of time and that's where something like Archera's insurance where something like Archera's insurance where something like Archera's insurance comes in where you're like, "Hey, this comes in where you're like, "Hey, this comes in where you're like, "Hey, this project works. I'm going to keep it up project works. I'm going to keep it up project works. I'm going to keep it up for a quarter. I don't want to be paying for a quarter. I don't want to be paying for a quarter. I don't want to be paying the crazy premium. I might even keep the crazy premium. I might even keep the crazy premium. I might even keep running it after the quarter, but I running it after the quarter, but I running it after the quarter, but I don't want to pay for the whole year. I don't want to pay for the whole year. I don't want to pay for the whole year. I don't feel comfortable there." And don't feel comfortable there." And don't feel comfortable there." And that's a great use case for something that's a great use case for something that's a great use case for something like an insurance provider like us to like an insurance provider like us to like an insurance provider like us to come in and say, "Hey, we'll share some come in and say, "Hey, we'll share some come in and say, "Hey, we'll share some of that risk with you." Uh but by of that risk with you." Uh but by of that risk with you." Uh but by default, you're going to get the best default, you're going to get the best default, you're going to get the best rate. And one of the really cool things rate. And one of the really cool things rate. And one of the really cool things about Archera is that if you get about Archera is that if you get about Archera is that if you get confidence that, "Hey, I'm going to use confidence that, "Hey, I'm going to use confidence that, "Hey, I'm going to use this to the rest of the term," you can this to the rest of the term," you can this to the rest of the term," you can just cancel the insurance and stop just cancel the insurance and stop just cancel the insurance and stop paying. Uh so, you're you're never

paying. Uh so, you're you're never paying. Uh so, you're you're never really locked in. You're always able to really locked in. You're always able to really locked in. You're always able to go back to what you're able to do kind go back to what you're able to do kind go back to what you're able to do kind of by default with the cloud provider. of by default with the cloud provider. of by default with the cloud provider. Um but you have now a partner that can Um but you have now a partner that can Um but you have now a partner that can offload that risk and allow you to kind offload that risk and allow you to kind offload that risk and allow you to kind of crawl, walk, run into those bigger of crawl, walk, run into those bigger of crawl, walk, run into those bigger commitments as the project matures commitments as the project matures commitments as the project matures versus being stuck at that promotional versus being stuck at that promotional versus being stuck at that promotional pricing where where you're not really pricing where where you're not really pricing where where you're not really going to be able to get a lot of going to be able to get a lot of going to be able to get a lot of leverage on your dollar. leverage on your dollar. leverage on your dollar. >> Yeah. >> Yeah. >> Yeah. Yeah, it feels like developers love Yeah, it feels like developers love Yeah, it feels like developers love elasticity. I definitely do. But like a elasticity. I definitely do. But like a elasticity. I definitely do. But like a CFO, they love predictability. They CFO, they love predictability. They CFO, they love predictability. They don't care what the developers think. don't care what the developers think. don't care what the developers think. And it feels like the modern cloud is And it feels like the modern cloud is And it feels like the modern cloud is arguably trying to sell to both sides. arguably trying to sell to both sides. arguably trying to sell to both sides. And it's something that can't fully And it's something that can't fully And it's something that can't fully coexist. And you are trying to make it coexist. And you are trying to make it coexist. And you are trying to make it coexist. You're trying to make both coexist. You're trying to make both coexist. You're trying to make both sides happy. sides happy. sides happy. >> Yeah, and and that's FinOps broadly, but >> Yeah, and and that's FinOps broadly, but >> Yeah, and and that's FinOps broadly, but I think in our space specifically of I think in our space specifically of I think in our space specifically of FinOps where we deal with, you know, FinOps where we deal with, you know, FinOps where we deal with, you know, rate optimization and commitments and rate optimization and commitments and rate optimization and commitments and upfront spend, we're very much at the upfront spend, we're very much at the upfront spend, we're very much at the forefront of that interface between, you forefront of that interface between, you forefront of that interface between, you know, the DevOps team, the know, the DevOps team, the know, the DevOps team, the infrastructure team, the office of the infrastructure team, the office of the infrastructure team, the office of the CIO or CTO, and the VP of finance, the CIO or CTO, and the VP of finance, the CIO or CTO, and the VP of finance, the FinOps team, the, you know, VP or the FinOps team, the, you know, VP or the FinOps team, the, you know, VP or the the CFO, right? And the CFO, right? And the CFO, right? And it is nuanced because they want to see it is nuanced because they want to see it is nuanced because they want to see things and understand things in very things and understand things in very things and understand things in very different ways, but ultimately they have different ways, but ultimately they have different ways, but ultimately they have the same goal, which is to get the best the same goal, which is to get the best the same goal, which is to get the best leverage on the dollar and the most leverage on the dollar and the most leverage on the dollar and the most predictability in a fundamentally predictability in a fundamentally predictability in a fundamentally unpredictable space. And so, that's what unpredictable space. And so, that's what unpredictable space. And so, that's what we're trying to do is trying to give we're trying to do is trying to give we're trying to do is trying to give them guardrails and bounds around what them guardrails and bounds around what them guardrails and bounds around what they are, you know, going to use and they are, you know, going to use and they are, you know, going to use and what they're going to spend, what they're going to spend, what they're going to spend, but also provide, again, we have a

but also provide, again, we have a but also provide, again, we have a completely free tool that comes along completely free tool that comes along completely free tool that comes along with, you know, the underwriting and all with, you know, the underwriting and all with, you know, the underwriting and all that we provide the options that we we that we provide the options that we we that we provide the options that we we give customers they can opt into give customers they can opt into give customers they can opt into to give them some of that additional to give them some of that additional to give them some of that additional visibility. And we have a ton of visibility. And we have a ton of visibility. And we have a ton of partners that help us with that as well partners that help us with that as well partners that help us with that as well because I think without that visibility, because I think without that visibility, because I think without that visibility, which we partner heavily on and give a which we partner heavily on and give a which we partner heavily on and give a lot of tools away for free on, you're lot of tools away for free on, you're lot of tools away for free on, you're not going to find alignment. And that not going to find alignment. And that not going to find alignment. And that alignment then speaks to the core of the alignment then speaks to the core of the alignment then speaks to the core of the problem on like, "Hey, there's a problem on like, "Hey, there's a problem on like, "Hey, there's a principal agent problem here. The DevOps principal agent problem here. The DevOps principal agent problem here. The DevOps person doesn't want to take risk. The person doesn't want to take risk. The person doesn't want to take risk. The finance person wants predictability and finance person wants predictability and finance person wants predictability and wants to actually put some risk and some wants to actually put some risk and some wants to actually put some risk and some commitment in place that they can commitment in place that they can commitment in place that they can actually track against. How do we actually track against. How do we actually track against. How do we reconcile that? Without that visibility, reconcile that? Without that visibility, reconcile that? Without that visibility, you never get that conversation, and you never get that conversation, and you never get that conversation, and then a solution like this never becomes, then a solution like this never becomes, then a solution like this never becomes, you know, clearly the go forward." you know, clearly the go forward." you know, clearly the go forward." >> Yeah. You know, you would think from >> Yeah. You know, you would think from >> Yeah. You know, you would think from talking to cloud folks who are just out talking to cloud folks who are just out talking to cloud folks who are just out there buying GPUs and building new data there buying GPUs and building new data there buying GPUs and building new data centers that everything is maxed out at centers that everything is maxed out at centers that everything is maxed out at 100%, and the only problem in the cloud 100%, and the only problem in the cloud 100%, and the only problem in the cloud is like build more spaces in a field in is like build more spaces in a field in is like build more spaces in a field in Utah somewhere. But it feels to me like Utah somewhere. But it feels to me like Utah somewhere. But it feels to me like the the the thing that no one's talking the the the thing that no one's talking the the the thing that no one's talking about underneath this is that there about underneath this is that there about underneath this is that there there is unused capacity and they need there is unused capacity and they need there is unused capacity and they need to find a way to to sell that and it to find a way to to sell that and it to find a way to to sell that and it seems like seems like seems like selling commitments is the way to solve selling commitments is the way to solve selling commitments is the way to solve that. I I wonder what's true.

that. I I wonder what's true. that. I I wonder what's true. >> I think it's it's a textured problem >> I think it's it's a textured problem >> I think it's it's a textured problem because for certain SKUs like take, you because for certain SKUs like take, you because for certain SKUs like take, you know, the Grace Blackwell 300s, right? know, the Grace Blackwell 300s, right? know, the Grace Blackwell 300s, right? Uh those are very [snorts] new Nvidia Uh those are very [snorts] new Nvidia Uh those are very [snorts] new Nvidia machines, you know, tough to deploy machines, you know, tough to deploy machines, you know, tough to deploy because they require liquid cooling and because they require liquid cooling and because they require liquid cooling and a lot of power. a lot of power. a lot of power. There's a clear supply demand imbalance There's a clear supply demand imbalance There's a clear supply demand imbalance there. You know, more recently, there. You know, more recently, there. You know, more recently, you know, you look at things like ARM you know, you look at things like ARM you know, you look at things like ARM CPUs in the data center. CPUs in the data center. CPUs in the data center. There's still plenty of those out there. There's still plenty of those out there. There's still plenty of those out there. In fact, for the last 2 years over 50% In fact, for the last 2 years over 50% In fact, for the last 2 years over 50% of the new CPUs in the data center of the new CPUs in the data center of the new CPUs in the data center on the AWS fleet were were Graviton and on the AWS fleet were were Graviton and on the AWS fleet were were Graviton and ARM-based. ARM-based. ARM-based. Um but, you know, you look at something Um but, you know, you look at something Um but, you know, you look at something like an x86 Intel CPU in the Azure data like an x86 Intel CPU in the Azure data like an x86 Intel CPU in the Azure data center, at least in a popular region, center, at least in a popular region, center, at least in a popular region, those might be harder to get. And so, those might be harder to get. And so, those might be harder to get. And so, again, it's a very textured problem in again, it's a very textured problem in again, it's a very textured problem in terms of like different SKUs have terms of like different SKUs have terms of like different SKUs have different demand supply profiles, but different demand supply profiles, but different demand supply profiles, but generally, I think in the areas of the generally, I think in the areas of the generally, I think in the areas of the market where there is a high degree of market where there is a high degree of market where there is a high degree of demand relative to supply, the demand relative to supply, the demand relative to supply, the commitment is the way that that problem commitment is the way that that problem commitment is the way that that problem is being solved. Um it's not a clean is being solved. Um it's not a clean is being solved. Um it's not a clean solve, obviously, because, you know, solve, obviously, because, you know, solve, obviously, because, you know, people want to make sure that if there's people want to make sure that if there's people want to make sure that if there's high demand, they can get the capacity high demand, they can get the capacity high demand, they can get the capacity um and they might not use it. They might um and they might not use it. They might um and they might not use it. They might just sit on it, but at least they're just sit on it, but at least they're just sit on it, but at least they're renting it out versus it sitting idle renting it out versus it sitting idle renting it out versus it sitting idle and not being rented in the data center.

and not being rented in the data center. and not being rented in the data center. And again, it's it's obviously a very And again, it's it's obviously a very And again, it's it's obviously a very dynamic market, but I would say that dynamic market, but I would say that dynamic market, but I would say that that's been the primary tool and when we that's been the primary tool and when we that's been the primary tool and when we talk to customers, we work with a ton of talk to customers, we work with a ton of talk to customers, we work with a ton of startups um that are running on the startups um that are running on the startups um that are running on the hyperscalers and even the neo clouds, hyperscalers and even the neo clouds, hyperscalers and even the neo clouds, you know, this is kind of the biggest you know, this is kind of the biggest you know, this is kind of the biggest hand-wringing for them. How much do they hand-wringing for them. How much do they hand-wringing for them. How much do they over commit so that they're ensuring over commit so that they're ensuring over commit so that they're ensuring their capacity and having again a their capacity and having again a their capacity and having again a partner like us in the mix where we can partner like us in the mix where we can partner like us in the mix where we can say, "Look, you know, you want to over say, "Look, you know, you want to over say, "Look, you know, you want to over commit 5x, we're happy to take 1x or 2x commit 5x, we're happy to take 1x or 2x commit 5x, we're happy to take 1x or 2x of that uh and risk share that with you of that uh and risk share that with you of that uh and risk share that with you for a price." for a price." for a price." Um it it gives them some of that Um it it gives them some of that Um it it gives them some of that flexibility. flexibility. flexibility. >> It's so funny how the >> It's so funny how the >> It's so funny how the things that I naively would think are things that I naively would think are things that I naively would think are intuitive are intuitive are intuitive are are not when you look at things at are not when you look at things at are not when you look at things at scale. I'm thinking about you said the scale. I'm thinking about you said the scale. I'm thinking about you said the word over commit and I'm thinking about word over commit and I'm thinking about word over commit and I'm thinking about times that I've been on a plane and times that I've been on a plane and times that I've been on a plane and they've told me they overbooked or the they've told me they overbooked or the they've told me they overbooked or the times that you've rented a car and you times that you've rented a car and you times that you've rented a car and you do that Jerry Seinfeld thing where it's do that Jerry Seinfeld thing where it's do that Jerry Seinfeld thing where it's like I have a reservation and they're like I have a reservation and they're like I have a reservation and they're like, well, no. like, well, no. like, well, no. Like, no, you should have the car here. Like, no, you should have the car here. Like, no, you should have the car here. I reserved it. I should have my finger I reserved it. I should have my finger I reserved it. I should have my finger on the car. Like, if they nail it, on the car. Like, if they nail it, on the car. Like, if they nail it, they're going to reserve everything.

they're going to reserve everything. they're going to reserve everything. They're going to be at 100% or 90% CPU They're going to be at 100% or 90% CPU They're going to be at 100% or 90% CPU all the time and everybody wins. all the time and everybody wins. all the time and everybody wins. But if they do it wrong, then it's a But if they do it wrong, then it's a But if they do it wrong, then it's a problem. So it feels like right sizing problem. So it feels like right sizing problem. So it feels like right sizing is it a technical problem is it a technical problem is it a technical problem or is it a organizational problem? Like or is it a organizational problem? Like or is it a organizational problem? Like is cloud waste technical waste or is it is cloud waste technical waste or is it is cloud waste technical waste or is it organizational waste? organizational waste? organizational waste? >> It's really a combined problem and I >> It's really a combined problem and I >> It's really a combined problem and I think the way that we think about it is think the way that we think about it is think the way that we think about it is there's really two sides to the cloud there's really two sides to the cloud there's really two sides to the cloud cost kind of world. There's the usage cost kind of world. There's the usage cost kind of world. There's the usage optimization which is saying, hey, it's optimization which is saying, hey, it's optimization which is saying, hey, it's how I'm running the most efficient way. how I'm running the most efficient way. how I'm running the most efficient way. Let's say I'm like serving, you know, Let's say I'm like serving, you know, Let's say I'm like serving, you know, Netflix videos. Is my cost per video Netflix videos. Is my cost per video Netflix videos. Is my cost per video served as efficient as possible if even served as efficient as possible if even served as efficient as possible if even if I just take the on-demand, you know, if I just take the on-demand, you know, if I just take the on-demand, you know, pay-as-you-go rate. Am I architected in pay-as-you-go rate. Am I architected in pay-as-you-go rate. Am I architected in an efficient way? an efficient way? an efficient way? You know, that's very much an You know, that's very much an You know, that's very much an engineering question and engineering engineering question and engineering engineering question and engineering problem. On the other side, there's rate problem. On the other side, there's rate problem. On the other side, there's rate optimization which is distinct from optimization which is distinct from optimization which is distinct from usage optimization and that's kind of usage optimization and that's kind of usage optimization and that's kind of where we play which is really what have where we play which is really what have where we play which is really what have I spent in the past, what can I I spent in the past, what can I I spent in the past, what can I comfortably commit to going forward and comfortably commit to going forward and comfortably commit to going forward and what is that stack of commitments end up what is that stack of commitments end up what is that stack of commitments end up looking like over time. And there's more looking like over time. And there's more looking like over time. And there's more fundamental questions there than just fundamental questions there than just fundamental questions there than just the infrastructure questions. It's like, the infrastructure questions. It's like, the infrastructure questions. It's like, how fast do I think my business is going how fast do I think my business is going how fast do I think my business is going to grow? Do I think I'm going to see to grow? Do I think I'm going to see to grow? Do I think I'm going to see customer churn? What does that mean in customer churn? What does that mean in customer churn? What does that mean in terms of what I'm comfortable committing terms of what I'm comfortable committing terms of what I'm comfortable committing to? In fact, even looking at that first to? In fact, even looking at that first to? In fact, even looking at that first world of usage optimization, you might world of usage optimization, you might world of usage optimization, you might say, hey, I have the ability to save 30% say, hey, I have the ability to save 30% say, hey, I have the ability to save 30% on my cluster. It's going to take a on my cluster. It's going to take a on my cluster. It's going to take a two-year migration, but I have that two-year migration, but I have that two-year migration, but I have that ability. Do I want to factor that in to ability. Do I want to factor that in to ability. Do I want to factor that in to what I commit? So, you know, visibility, what I commit? So, you know, visibility, what I commit? So, you know, visibility, forecasting, and really planning that forecasting, and really planning that forecasting, and really planning that commitment portfolio is in that second commitment portfolio is in that second commitment portfolio is in that second bucket. And And again, that's more of a

bucket. And And again, that's more of a bucket. And And again, that's more of a finance problem, but they're very finance problem, but they're very finance problem, but they're very related, right? These things, again, related, right? These things, again, related, right? These things, again, have to speak to each other and use some have to speak to each other and use some have to speak to each other and use some sort of common language, as I was saying sort of common language, as I was saying sort of common language, as I was saying earlier. earlier. earlier. >> Right. Now, you said that you have >> Right. Now, you said that you have >> Right. Now, you said that you have petabytes of data, like you're getting petabytes of data, like you're getting petabytes of data, like you're getting aggregated data across, you know, kind aggregated data across, you know, kind aggregated data across, you know, kind of anonymized aggregated data across all of anonymized aggregated data across all of anonymized aggregated data across all of it, and that's how you're able to do of it, and that's how you're able to do of it, and that's how you're able to do your pricing. But you would think that your pricing. But you would think that your pricing. But you would think that companies would have their own usage companies would have their own usage companies would have their own usage data. So, why would a company overbuy or data. So, why would a company overbuy or data. So, why would a company overbuy or underbuy underbuy underbuy when they themselves have years of usage when they themselves have years of usage when they themselves have years of usage data? Is that a blind spot on their data? Is that a blind spot on their data? Is that a blind spot on their part? part? part? >> Well, the thing is that, you know, >> Well, the thing is that, you know, >> Well, the thing is that, you know, there's kind of two aspects to this. One there's kind of two aspects to this. One there's kind of two aspects to this. One is just understanding that usage data in is just understanding that usage data in is just understanding that usage data in and of itself is difficult. That's why and of itself is difficult. That's why and of itself is difficult. That's why there's this whole vendor ecosystem of there's this whole vendor ecosystem of there's this whole vendor ecosystem of folks to help which is getting folks to help which is getting folks to help which is getting visibility on cloud costs. Many of our visibility on cloud costs. Many of our visibility on cloud costs. Many of our partners actually kind of play in that partners actually kind of play in that partners actually kind of play in that space as an example. And so, just the space as an example. And so, just the space as an example. And so, just the expertise to understand, distill, and expertise to understand, distill, and expertise to understand, distill, and then forecast based on that data is then forecast based on that data is then forecast based on that data is non-trivial. But even beyond that, there non-trivial. But even beyond that, there non-trivial. But even beyond that, there are known unknowns. So, even if you have are known unknowns. So, even if you have are known unknowns. So, even if you have perfect forecasting in your data, known perfect forecasting in your data, known perfect forecasting in your data, known unknowns like how fast will I grow my unknowns like how fast will I grow my unknowns like how fast will I grow my business? If I am just towing that line business? If I am just towing that line business? If I am just towing that line and and committing, you know, to every and and committing, you know, to every and and committing, you know, to every marginal GPU as a customer comes on, marginal GPU as a customer comes on, marginal GPU as a customer comes on, what happens if we have a capacity what happens if we have a capacity what happens if we have a capacity crunch and I'm not able to get more GPUs crunch and I'm not able to get more GPUs crunch and I'm not able to get more GPUs to serve the marginal customers? Should to serve the marginal customers? Should to serve the marginal customers? Should I overcommit now? That's something that I overcommit now? That's something that I overcommit now? That's something that actually a lot of our startups are actually a lot of our startups are actually a lot of our startups are thinking about, like do I just buy a lot thinking about, like do I just buy a lot thinking about, like do I just buy a lot more now, a bigger chunk, and grow into more now, a bigger chunk, and grow into more now, a bigger chunk, and grow into it and not take the risk that I won't it and not take the risk that I won't it and not take the risk that I won't have capacity down the line. You know, have capacity down the line. You know, have capacity down the line. You know, these are kind of the known unknowns. I these are kind of the known unknowns. I these are kind of the known unknowns. I mean, will my architecture change? Even mean, will my architecture change? Even mean, will my architecture change? Even if you are the person doing the usage if you are the person doing the usage if you are the person doing the usage optimization, you don't know if your

optimization, you don't know if your optimization, you don't know if your your database architecture is going to your database architecture is going to your database architecture is going to look the same in in 18 months. No one look the same in in 18 months. No one look the same in in 18 months. No one does. So, it's really those known does. So, it's really those known does. So, it's really those known unknowns that having a partner like us unknowns that having a partner like us unknowns that having a partner like us can come in and help because we've seen can come in and help because we've seen can come in and help because we've seen now thousands of customers that look now thousands of customers that look now thousands of customers that look like you. So, you might only see your like you. So, you might only see your like you. So, you might only see your own data. We've seen own data. We've seen own data. We've seen tons of customers that look like you's tons of customers that look like you's tons of customers that look like you's data. So, we actually have a prior. data. So, we actually have a prior. data. So, we actually have a prior. We're able to price that in a sense more We're able to price that in a sense more We're able to price that in a sense more effectively effectively effectively than you might be. than you might be. than you might be. >> Okay, that's funny that you say that >> Okay, that's funny that you say that >> Okay, that's funny that you say that because that's kind of where my next because that's kind of where my next because that's kind of where my next question was kind of going, which is question was kind of going, which is question was kind of going, which is like all of this sounds simple, but like all of this sounds simple, but like all of this sounds simple, but pricing risk by its definition is hard. pricing risk by its definition is hard. pricing risk by its definition is hard. So, I was going to ask what are the So, I was going to ask what are the So, I was going to ask what are the signals signals signals that would allow our chair to say, "This that would allow our chair to say, "This that would allow our chair to say, "This is a workload that is safe enough to is a workload that is safe enough to is a workload that is safe enough to insure." insure." insure." That's data. That's data. That's data. >> Yes. >> Yes. >> Yes. Well, there's there's a lot, right? So, Well, there's there's a lot, right? So, Well, there's there's a lot, right? So, it's what is the specific and just to it's what is the specific and just to it's what is the specific and just to give you a sense, we give you a sense, we give you a sense, we are are pretty lightweight. We'll get to are are pretty lightweight. We'll get to are are pretty lightweight. We'll get to your to your point kind of anonymize your to your point kind of anonymize your to your point kind of anonymize metadata on what is the spend, what is metadata on what is the spend, what is metadata on what is the spend, what is up, what is down, you know, if there up, what is down, you know, if there up, what is down, you know, if there tags on these machines like production tags on these machines like production tags on these machines like production versus R&D.

versus R&D. versus R&D. Basically, just the metadata around the Basically, just the metadata around the Basically, just the metadata around the billing and usage in the environment. billing and usage in the environment. billing and usage in the environment. But even through that, you know, just But even through that, you know, just But even through that, you know, just what are the machines being used? What what are the machines being used? What what are the machines being used? What is the type of business that this is, is the type of business that this is, is the type of business that this is, right? just a simple Crunchbase look up right? just a simple Crunchbase look up right? just a simple Crunchbase look up or looking in our CRM on how understand or looking in our CRM on how understand or looking in our CRM on how understand the employees and funding, understanding the employees and funding, understanding the employees and funding, understanding the tags on the machines, understanding the tags on the machines, understanding the tags on the machines, understanding you know, things like very specific, you you know, things like very specific, you you know, things like very specific, you know, services and know, services and know, services and their lifetime. So, if someone is their lifetime. So, if someone is their lifetime. So, if someone is migrating and on something like, you migrating and on something like, you migrating and on something like, you know, Azure VMware service, they're very know, Azure VMware service, they're very know, Azure VMware service, they're very likely to start migrating some of that likely to start migrating some of that likely to start migrating some of that capacity to AKS within a year. Okay, so capacity to AKS within a year. Okay, so capacity to AKS within a year. Okay, so I probably won't offer that much on the I probably won't offer that much on the I probably won't offer that much on the you know, Azure VMware service you know, Azure VMware service you know, Azure VMware service commitments relative to something like commitments relative to something like commitments relative to something like AKS. So, there's a lot of really AKS. So, there's a lot of really AKS. So, there's a lot of really granular decisioning and you know, the granular decisioning and you know, the granular decisioning and you know, the beautiful thing about modern machine beautiful thing about modern machine beautiful thing about modern machine learning is that you can basically put learning is that you can basically put learning is that you can basically put this all into a feature vector and have this all into a feature vector and have this all into a feature vector and have you know, you know, you know, a set of algorithms basically rank these a set of algorithms basically rank these a set of algorithms basically rank these things for you you know, from random things for you you know, from random things for you you know, from random forest through to deep neural networks forest through to deep neural networks forest through to deep neural networks and actually come up with some of these and actually come up with some of these and actually come up with some of these statistical correlations and outcomes of statistical correlations and outcomes of statistical correlations and outcomes of the historical data that we're able to the historical data that we're able to the historical data that we're able to apply in a future-looking way.

apply in a future-looking way. apply in a future-looking way. >> Now, with all the data you have, is it >> Now, with all the data you have, is it >> Now, with all the data you have, is it all people with insurance? Or because all people with insurance? Or because all people with insurance? Or because would wouldn't insurance by definition would wouldn't insurance by definition would wouldn't insurance by definition change customers behavior? change customers behavior? change customers behavior? And someone would take a different risk And someone would take a different risk And someone would take a different risk if the downside is bounded. if the downside is bounded. if the downside is bounded. >> Yeah, actually and it's a great >> Yeah, actually and it's a great >> Yeah, actually and it's a great question. It speaks to that moral hazard question. It speaks to that moral hazard question. It speaks to that moral hazard point. So, so two things. One is we point. So, so two things. One is we point. So, so two things. One is we actually don't see actually don't see actually don't see too much moral hazard because of some of too much moral hazard because of some of too much moral hazard because of some of the things like you have to run the the things like you have to run the the things like you have to run the workload for a certain amount of time workload for a certain amount of time workload for a certain amount of time beforehand before we actually will offer beforehand before we actually will offer beforehand before we actually will offer insurance. Things like that. But but insurance. Things like that. But but insurance. Things like that. But but even that aside, only about 40% of the even that aside, only about 40% of the even that aside, only about 40% of the customers on our platform at any time customers on our platform at any time customers on our platform at any time are procuring insurance from us. They are procuring insurance from us. They are procuring insurance from us. They love to have love to have love to have you know, one we offer a lot of free you know, one we offer a lot of free you know, one we offer a lot of free services and tools to understand your services and tools to understand your services and tools to understand your bill, to do some of the basic stuff bill, to do some of the basic stuff bill, to do some of the basic stuff around forecasting and commitment around forecasting and commitment around forecasting and commitment planning. We'll actually automate planning. We'll actually automate planning. We'll actually automate native commitment life cycle management native commitment life cycle management native commitment life cycle management for customers entirely for free. for customers entirely for free. for customers entirely for free. You know, and a lot of customers will You know, and a lot of customers will You know, and a lot of customers will use us for that. use us for that. use us for that. Really our goal is to put these quotes Really our goal is to put these quotes Really our goal is to put these quotes for you know, if and when insurance for you know, if and when insurance for you know, if and when insurance makes sense for both parties, for us and makes sense for both parties, for us and makes sense for both parties, for us and them, in front of the customers in in a them, in front of the customers in in a them, in front of the customers in in a timely manner, like in the right way.

timely manner, like in the right way. timely manner, like in the right way. And so, as a function of that, you know, And so, as a function of that, you know, And so, as a function of that, you know, we see a good amount of uninsured we see a good amount of uninsured we see a good amount of uninsured committed spend even from customers committed spend even from customers committed spend even from customers buying insurance from us. buying insurance from us. buying insurance from us. But we see a lot of customers that that But we see a lot of customers that that But we see a lot of customers that that don't have that spend that are, you don't have that spend that are, you don't have that spend that are, you know, maybe waiting just for the right know, maybe waiting just for the right know, maybe waiting just for the right quote from us to go and buy it at some quote from us to go and buy it at some quote from us to go and buy it at some point. All right, that's the hope. point. All right, that's the hope. point. All right, that's the hope. >> This will this will might be putting you >> This will this will might be putting you >> This will this will might be putting you on the spot, but you know, cloud bills on the spot, but you know, cloud bills on the spot, but you know, cloud bills are notoriously weird. Is what is the are notoriously weird. Is what is the are notoriously weird. Is what is the gnarliest billing edge case that you can gnarliest billing edge case that you can gnarliest billing edge case that you can think of, you know, keeping anonymous? think of, you know, keeping anonymous? think of, you know, keeping anonymous? Like what's the weirdest thing you've Like what's the weirdest thing you've Like what's the weirdest thing you've bumped into in your experience? bumped into in your experience? bumped into in your experience? >> Yeah, without talking about any specific >> Yeah, without talking about any specific >> Yeah, without talking about any specific customers, customers, customers, once you start getting to the scale of once you start getting to the scale of once you start getting to the scale of like 80, 90 million dollars a year with like 80, 90 million dollars a year with like 80, 90 million dollars a year with a single provider, there are crazy a single provider, there are crazy a single provider, there are crazy things that come up. Like first of all, things that come up. Like first of all, things that come up. Like first of all, you have to upgrade your database just you have to upgrade your database just you have to upgrade your database just to ingest the bill and parse it. But but to ingest the bill and parse it. But but to ingest the bill and parse it. But but two, you'll actually see that on, you two, you'll actually see that on, you two, you'll actually see that on, you know, every bill or every other bill know, every bill or every other bill know, every bill or every other bill depending on the scale, the provider depending on the scale, the provider depending on the scale, the provider themselves will often have billing themselves will often have billing themselves will often have billing errors in putting it together, errors in putting it together, errors in putting it together, especially because you have custom especially because you have custom especially because you have custom pricing and other things, you know, at pricing and other things, you know, at pricing and other things, you know, at that scale. I think the other thing, you that scale. I think the other thing, you that scale. I think the other thing, you know, especially when you look at things know, especially when you look at things know, especially when you look at things like the Azure ecosystem and how they've like the Azure ecosystem and how they've like the Azure ecosystem and how they've changed their contracting structure from changed their contracting structure from changed their contracting structure from like enterprise agreements now to maxes, like enterprise agreements now to maxes, like enterprise agreements now to maxes, there's so much being lost in there's so much being lost in there's so much being lost in translation there, especially for the translation there, especially for the translation there, especially for the big scaled accounts that big scaled accounts that big scaled accounts that um the bills themselves are often not um the bills themselves are often not um the bills themselves are often not treatable as a source of truth. And I treatable as a source of truth. And I treatable as a source of truth. And I think once that starts to break, a lot think once that starts to break, a lot think once that starts to break, a lot of things downstream of that, you know, of things downstream of that, you know, of things downstream of that, you know, uh start to break. So So those have been uh start to break. So So those have been uh start to break. So So those have been kind of the class of problems that I kind of the class of problems that I kind of the class of problems that I think have been the most gnarly and think have been the most gnarly and think have been the most gnarly and interesting when you can't trust the interesting when you can't trust the interesting when you can't trust the bill itself coming from the provider.

bill itself coming from the provider. bill itself coming from the provider. >> Wow. >> Wow. >> Wow. Okay. So for folks that are listening, Okay. So for folks that are listening, Okay. So for folks that are listening, what is the right size when they think, what is the right size when they think, what is the right size when they think, you know, I think we're big enough we you know, I think we're big enough we you know, I think we're big enough we need cloud insurance. We're startup, need cloud insurance. We're startup, need cloud insurance. We're startup, we've we're make we're selling this much we've we're make we're selling this much we've we're make we're selling this much or we're spending that much. Maybe I, or we're spending that much. Maybe I, or we're spending that much. Maybe I, Scott Hanson, don't need it because I'm Scott Hanson, don't need it because I'm Scott Hanson, don't need it because I'm little or maybe I do. What's the right little or maybe I do. What's the right little or maybe I do. What's the right size for someone to reach out? size for someone to reach out? size for someone to reach out? >> Yeah, well, frankly, we work with >> Yeah, well, frankly, we work with >> Yeah, well, frankly, we work with customers that, you know, range from a customers that, you know, range from a customers that, you know, range from a two-person startup on credits where two-person startup on credits where two-person startup on credits where they're making their first commitment they're making their first commitment they're making their first commitment ever to like a database that they ever to like a database that they ever to like a database that they want to save like 50 bucks on. And want to save like 50 bucks on. And want to save like 50 bucks on. And again, it's very lightweight, there's no again, it's very lightweight, there's no again, it's very lightweight, there's no infrastructure changes, 5 minutes to infrastructure changes, 5 minutes to infrastructure changes, 5 minutes to share your bill and get a quote. Um so share your bill and get a quote. Um so share your bill and get a quote. Um so so again, we work with like small YC so again, we work with like small YC so again, we work with like small YC companies like that. But generally, what companies like that. But generally, what companies like that. But generally, what we say is when you're around, you know, we say is when you're around, you know, we say is when you're around, you know, 100 to 200,000 dollars a year of spend 100 to 200,000 dollars a year of spend 100 to 200,000 dollars a year of spend in the cloud, that's when the numbers in the cloud, that's when the numbers in the cloud, that's when the numbers start to become very meaningful. Like start to become very meaningful. Like start to become very meaningful. Like we're talking about, you know, funding we're talking about, you know, funding we're talking about, you know, funding an initiative, like a whole new AI an initiative, like a whole new AI an initiative, like a whole new AI initiative or potentially we're even initiative or potentially we're even initiative or potentially we're even talking about funding a headcount once talking about funding a headcount once talking about funding a headcount once we get to the 500k we get to the 500k we get to the 500k plus a year plus a year plus a year level. And so, you know, I would say level. And so, you know, I would say level. And so, you know, I would say that really people use us that really people use us that really people use us as an individual developer tool, small as an individual developer tool, small as an individual developer tool, small startups use us, startups use us, startups use us, especially now with just being required especially now with just being required especially now with just being required to get a commitment to sign a GPU. If to get a commitment to sign a GPU. If to get a commitment to sign a GPU. If there's two guys in a garage and want a there's two guys in a garage and want a there's two guys in a garage and want a GPU, they'll often use us for that.

GPU, they'll often use us for that. GPU, they'll often use us for that. Um but but the meaningful numbers start Um but but the meaningful numbers start Um but but the meaningful numbers start to come at that 100k to 500k scale in to come at that 100k to 500k scale in to come at that 100k to 500k scale in terms of to a business to an enterprise. terms of to a business to an enterprise. terms of to a business to an enterprise. >> Okay. That's that's that's more >> Okay. That's that's that's more >> Okay. That's that's that's more accessible than I realized. I I accessible than I realized. I I accessible than I realized. I I a a a startup and a couple of folks in a a a a startup and a couple of folks in a a a a startup and a couple of folks in a garage really could potentially save garage really could potentially save garage really could potentially save some money on that. some money on that. some money on that. >> Yeah, and it's often meaningful to them. >> Yeah, and it's often meaningful to them. >> Yeah, and it's often meaningful to them. Like we have a guy who runs a side Like we have a guy who runs a side Like we have a guy who runs a side project that, you know, has scaled up to project that, you know, has scaled up to project that, you know, has scaled up to now 20k now 20k now 20k um per month, but it was running at $200 um per month, but it was running at $200 um per month, but it was running at $200 a month. And they bought some cloud a month. And they bought some cloud a month. And they bought some cloud insurance, ended up saving 75 bucks a insurance, ended up saving 75 bucks a insurance, ended up saving 75 bucks a month. It was meaningful for him, you month. It was meaningful for him, you month. It was meaningful for him, you know, running off of his credit card. know, running off of his credit card. know, running off of his credit card. >> Fantastic. All right, well, folks can >> Fantastic. All right, well, folks can >> Fantastic. All right, well, folks can learn more at Archera.ai, that's a r c h learn more at Archera.ai, that's a r c h learn more at Archera.ai, that's a r c h e r a.ai. And I have been chatting with e r a.ai. And I have been chatting with e r a.ai. And I have been chatting with Aron Khanna, co-founder and CEO of Aron Khanna, co-founder and CEO of Aron Khanna, co-founder and CEO of Archera. Thanks for hanging out with me Archera. Thanks for hanging out with me Archera. Thanks for hanging out with me today. today. today. >> Thanks so much for having me, Scott. It >> Thanks so much for having me, Scott. It >> Thanks so much for having me, Scott. It was great to chat. was great to chat. was great to chat. >> This has been another episode of >> This has been another episode of >> This has been another episode of Hanselminutes, and we'll see you again Hanselminutes, and we'll see you again Hanselminutes, and we'll see you again next week.