Aran Khanna, Archera & Steve Baer, Kemet Financial Technologies | AWS Startup Showcase S3 E2

hey everyone welcome to thecube's presentation of the AWS startup showcase analytics and cost optimization this is season three episode two of our ongoing series that covers very exciting startups from the AWS ecosystem I'm your host Lisa Martin today excited to be joined by one of our Cube Alum Aran Khanna the CEO and co-founder of archera and Steve Baer joins as well senior site reliability engineer at our chair as customer chemit trading guys great to have you here thank you so much for joining me on the program thank you so much for having us we're going to be talking about our chair of free automated Cloud management we're also going to be talking about chemet's trading's use case so Iran you and your co-founder brother nikkiel you guys were on thecube just a month or so ago talking with our co-founder John Furrier great story as you guys talked about with John there's a lot of ways to save on the cloud there's a lot of vendors that are saying to customers hey we can get you Cloud savings but our chair is different I know that you have a different model talk to me about that yeah absolutely so to go back a little bit to really the founding story of archera one of our key experiences from our team is actually a lot of us are former amazonians I myself was actually part of the launch of the sagemaker program at AWS which includes the sagemaker platform and a lot of the uh managed inference apis around things like recognition and Lex and poly and one of the big issues that we saw in customers adopting these you know as everyone has learned now very expensive Ai and ml Services based on deep learning was cost and you know what I kept finding with my customers was that while AWS and Cloud providers like Azure have native tools that were free or low cost to try and help them address some of these problems and adopt these new services in a cost-conscious way they were fairly hard to use not really integrated with each other and it actually cost a lot of engineering time to string these things together and on the other side there were a lot of vendors I would work with who were promising to accelerate the process of of putting a cost optimization in a fin Ops practice if you will in place but they would come into my customers and say hey we can save you X dollars but we won't tell you how until you lock into a few annual contracts or a strategy that obligates you to basically give up some ongoing percentage of spend and savings so it left a lot of my customers between a rock and a hard place and that's really what precipitated us starting our chair up we wanted to create a bit of a new model as you referenced with a free single platform that acts as a replacement for those native AWS and other Cloud providers free tools and fills the gaps in their offerings with a low floor high ceiling type of approach with you know the ability for a customer who really wants to get something in place but doesn't have the time to come in and click a button and basically get you know sixty percent of the way there from a savings perspective or really Advanced fin Ops teams to come in and dig down and figure out what are the individual resources and commitments they want to be uh covering and using going forward so that's really the model that we've pioneered and one of the key ways that we've tried to go to market is to be really customer obsessed there's no two-way there's no one-way doors it's all two-way doors that's actually one of the key Innovations we're bringing to this Market in filling the gaps in the AWS ecosystem particularly around commitments and there's never a hidden savings tax there's no surprise bills so we're really trying to Pioneer this customer obsessed model and one of the key things that we found that customers really like is the fact that we actually have a new solution that fills the gaps between on-demand one year and three-year commitments with a flexible offering that allows customers to commit to any aw service not just compute but things like sagemaker as well for as short as 30 days and then have the option at any point after that to click a button and send that commitment right off their books and that's been able to unlock a lot of new use cases for customers and the Beautiful Thing is customers can use our platform and opt into that and pay us money as part of a adopting those commitments or they can just use our platform to automate and plan native AWS commitments save millions of dollars and never pay us a penny and that's really the model that we're trying to bring into the ecosystem to be very customer obsessed very customer obsessed also very transparent it sounds like so you you did a great job Iran of talking about the catalysts that were there to start our chair the gaps that you're feeling in the market you talked about use cases we've got Steve here from chemit Trading we're going to be talking about his use case in a minute but Iran our chair of focus is on automating commitment management management excuse me can you dig into what that category of optimization really entails and what it is that our chair delivers in that category yeah absolutely so when you think about commitments at AWS there are as an example their commitments on all the different clouds but we'll talk about AWS because it has the the largest number of different commitment types there are 36 plus ways to commit to Any Given workload and depending on how long that workload will be up what the actual savings rate is and a number of different factors you actually need to navigate this as an end user and that's really really difficult to do especially if you're not an expert in uh you know Cloud pricing which of course no one wants to become an expert in they want to go and build software and deliver value to end users so what we do at archera is provide the tools to not only look at all the different commitment types and build an AI driven forecast of our customers workloads but then come up with a set of Blended default strategies blending all the different commitment types all the term lengths all the different contract types including our 30-day Gris our garage guaranteed buyback insured instances to really give customers a set of options that in one click they can apply to their infrastructure or they could actually dig in build a really customized automation strategy and uh deploy that so again that's that low floor and high ceiling approach but generally what we're trying to do is make sure that customers are able to instead of taking weeks to do this take hours or even minutes to do this no matter how deep down the stack they want to go and one of the other very unique things that we're doing is taking that next step after purchasing commitments where you're actually reselling them or exchanging them and making that entirely free to customers that's something that again a lot of customers are on things like convertible Reserve instances are used to paying you know some ungodly percentage of ongoing savings or spent they're coming to us and getting it for free so really trying to hone in on where we're providing value where we take a cost on our books with this guaranteed buyback and trying to give the rest of that Universe away to the customers for free to get them in to the motion of actually automating this process without that downside risk of having an ongoing tax on your total spend or savings you get to keep as much of that as makes sense for your organization well that the value prop is quite clear as well let's get dig into an actual use case we've got Steve bear here as we introduced in the beginning of the show from chemet Trading Steve welcome talk about what the company does and also tell us a little bit about your role I understand you are a jack of all trades sure so chemist trading is a financial technology provider to institutions on the buy side so if you're doing trading specifically for us we Service uh digital asset derivatives so a very Niche space and a nascent Market small team moving fast hopefully not breaking anything it's kind of the kind of the vibe we have going as you mentioned a title is senior SRE but that pretty much means everything at the platform on down so if it's you know dealing with colos to managing our whole Cloud infra to self-service platform engineering and then of course you know operational excellence and the SRE practices so lots to do there um on your plate go ahead sorry yes oh sure um so we found our chair through the AWS partner Network actually through our AWS rep through startup so it's a wonderful easy way to get access we got it installed through the marketplace dropped in some cloud formation and we were in as we're in kind of mentioned the barrier to entry the initial onboarding was easy and it was a no-brainer for us as you might imagine uh in general AWS billing is notorious for being difficult it was just to begin to understand even what was going on we had tons of Greenfield development moving rapidly and need to bring some uh some maturation to to how we were handling our spin what were some of the key technical requirements Steve that that you and Kevin had that when you saw our chair of the light just the light bulb went off what were some of those key capabilities that it had to deliver sure well I mean for us in the space that we work and being a early stage company we tried to you know balance as much as possible the uniqueness of the regulatory things that our customers and therefore us may have to deal with so the long-term commitment of a zonal or Regional RI was a non-starter for us and as everyone mentioned the gri really hit that sweet spot where we could begin to take advantage of the cost benefit of some of these commitments without any of the risk that you know uh some jurisdiction may change how things operate and we have to then move so talk a little bit more about the use case and are there are there some cool Iran did a great job of describing some of the significant cost benefits that organizations can get it also sounds like the time to Value as well what are some of the big outcomes that you guys are achieving so far sure well I mean we get initial price breaks on our ECS based workloads in fargate so just getting a baseline don't have to manage that don't have to deal with any of that TurnKey out out the box as we are driving that that product maturity as we are either moving to Dedicated ECS back to you know ec2 hosts we can then move to Gris for those so whether we move further into a cloud native Technologies or more into the host space more traditional VM based Technologies no matter which way we grow and mature our architecture we can keep cost at the front of our consideration which is critical Iran talk a little bit about what Steve shared in terms of the chemit trading use case I imagine there's a lot of similarities with what you see with other companies in different Industries yeah absolutely one of the key things that uh we try and put in place from a process perspective for all of our customers no matter the industry is to firstly try and get them a win up front because we offer this unique 30-day discount that really doesn't have any alternative in the AWS ecosystem it unlocks net new ways to commit to the cloud that means that as soon as we install with the customer within you know the first few hours we're able to actually click a button cover Resources with commitments and get them immediate wins from a savings perspective because there are usually workloads like seasonality or short-term projects that are just not being covered today with commitments that that 30-day gri that Steve was talking about unlocks and that's a pretty Universal thing but then as our customers mature and you know Steve is talking about the evolution of their architecture over time one of the other key things we do from a commitment perspective is help them actually build very granular long-term forecasts with what we call our scenario planner and that allows them not only to figure out all right how do I save money today by covering stuff that I'm uncertain about but then how do I actually plan out and get more certain about the future figure out what those future workloads could be and then start to evolve my commitment strategy so I can commit to you know more things for longer use Gris where it makes sense use native AWS commitments where it makes sense but overall find the lowest total cost of ownership for my workloads over the long term given my engineering plans and that's where we get uh with particularly some of our larger customers into things like helping them negotiate the Enterprise deal with their Cloud vendor like AWS or Azure and that actually encompasses not just thinking about the commitments but thinking about the infrastructure and that's where we start to work more with engineering teams now our platform's entirely free and we give a ton of visibility chargeback showback out of the box for free to those engineering teams they can see what they're using but additionally we have this scenario forecasting tool that allows Engineers before they spin things up or as they're are spinning things up and growing to model out what did that future spend going to be and contribute that Upstream into a real Global forecast for a customer that really drives what is that top level you know CIO level bill or check rather that they're going to write uh for the next one to three to five years to their cloud provider so really going from the granular resource commitment at the beginning of our engagement all the way to that top level Enterprise commitment as we kind of go through the circle and really help customers mature their fin Ops process so so really a very complete almost visibility tool that you're enabling organizations to be able to see from those prioritization of resources to the Top Line which is key you talked about Fin Ops Steve I know a little bit about chemit you described a little bit about it I'm running an extensive Cloud environment talk about from your seat what are some of the major drivers for building a successful finops practice in a company that is really fast growing and that is cloud native what Iran talked about getting you know in sync with the engineers what do engineers need to do to help make finnops successful oh man I mean just starting with your your spend Opex in mind from Day Zero we oftentimes think of cost Management in or cost control in these sort of restrictive ways as a optimization after the fact and really especially in the current macro being mindful of your of your spent as you're trying to grow and trying to move fast and deliver a successful product that is engaging with customers and meeting their expectations oh I can go wrong you know with the benefit of Greenfield development you know we employ service Technologies as much as possible I mentioned how far gate we might be moving some of those workloads to host based um but you know due to the nature of our workloads with dedicated hosts the overhead they bring it's kind of unavoidable and being quickly required Now default on demand but everything across the board you very quickly are spending far more than you would have otherwise and with something as simple as our chair in these offerings Gras really hit that sweet spot to allow us to make that consideration and take that consideration and take action on that consideration um from day one and that's the key really is to be able to take action on that Iran talk a little bit about from a cultural perspective bringing engineering folks together devops folks together who may not be as familiar with Cloud pricing and optimization or speak the language or to your point earlier really need should they be spending their resources there how does our chair help bridge the engineering folks the devops folks so that they can achieve those big savings that you're talking about yeah so one of the key things that we're trying to do with uh our devops user specifically is give them one a really really easy interface uh much simpler than you know the AWS cost Explorer to see just default what their individual infrastructure what their individual piece of the overall AWS Pi is costing uh on a near real-time basis of giving them smart alerting and smart reporting around that so I think really starts with the visibility piece that again we're providing for free for our customers uh but then it goes into kind of longer term things like custom dashboards where you can actually bring in uh the actual units that you're serving and look at unit economics and cost of goods sold uh as sort of the next level metric uh and we really allow Engineers to go through that whole process from just seeing what is being spent to understanding what are they getting for what is being spent and what is that kind of as the you know product goes into maturity and actually is being served to end customers what is that metric that they really want to optimize around because often if you're spending more it could be be a good thing because you're saving more users right so that's one of the key things from a visibility perspective we're trying to bring to engineers and we have a uh you know an Enterprise ready sort of bi Solution that's very customizable to really help customers go from the simple free product that we provide to something much more sophisticated and then on the other side we essentially will help customers on the engineering side integrate things like the scenario forecasting into their workflows and part of that key part of our scenario forecasting tool is helping customers understand on the devops side what is actually as I'm spinning something up what is actually the alternative that I have here is there a cheaper region could I be running on graviton or is you know AWS compute Optimizer saying that this thing is way too oversized that I should be running you know two two instant sizes smaller all of that data is in you know 10 50 in disparate places between the AWS simple pricing calculator and Trust advisor and compute Optimizer we bring it all into one place and increasingly are putting it into the engineering workflow where they're working in GitHub uh you know in their infrastructure as code environment Etc so that's really the way that we think about bringing that uh kind of disparate set of information at the fingertips of devops users and they always want to do the right thing the engineers the devops users so it's just about making it easy for them yeah really streamlining their productivity as well it sounds like Steve my last question for you is if you are talking to a prospect if Iran says hey Steve you got this this Hot Prospects they want to talk to a customer who's really successful with our technology what would you say if someone said Steve why are chera why Steve what do you think for us it was the easy onboarding and the no cost to get started I think it with the ease of the onboarding process getting up and just bringing visibility of these things the the value prop for the the initial input is is Extreme and that's even if it's just a free that's a free offering at that point so even if you aren't taking advantage of the whole Suite of things that our chair is bringing to the table just beginning to frame and have that single pane of glass that you can begin to interrogate these things throughout your cloud of state is it's critical so strong value prop and Stevie did a great job of really articulating our chairs value prop through the chemit trading use case Iran I want to leave you with the final words here what are some of the final kind of takeaways that you want our audience to know about archera the approach this unique approach that you're bringing to Cloud cost management well you know to Steve's point I think uh one of the things that we're pioneering is a true two-way door in the cost management space from a vendor perspective again we want to be really customer obsessed and the two things we wanted to do in starting this customer and this company was one reset essentially the price floor so it's no longer based on a percentage of savings or spend but it could be completely free for the customers to go and use native tools and then it gives them access to more advanced strategies like using Gris uh or our you know custom bi and Reporting dashboards that then they can opt in to pay a flat sort of SAS rate for based on our costs not based on their spender savings so that was sort of one piece of it the other piece was actually to uh really revolutionize in a way the way that people work instead of providing just visibility we want to provide automation we want to provide more integration with devops workflows and eventually have this 24 7 you know finops AI co-pilot that's really helping you not just by giving you great default plans when you log into the platform that you can go and play with an implementing The Click but really going deeper into the engineering workflow to help you and your teams choose the right resources up front and plan out for those long term three to five year Enterprise uh sort of road map discussions that you have with your Cloud vendor when you start getting to scale and you know I'd say that one of the key things that I want to leave folks with is just the echo what Steve said there's really no risk in trying us we're never going to charge a percentage of spend or savings we like to say we compete with AWS cost Explorer because we're a free tool and we're about as easy to turn on and use and you know that's the way that we think cost management uh for any platform should exist in the future and that's the world that we're trying to build for end customers awesome work guys revolutionize the way people work cost optimization no risk who could say no to that Iran Steve thank you so much for coming on thecube for the AWS startup showcase analytics and cost optimization we really enjoyed the time with you thank you thank you thank you my pleasure we want to thank you for watching and remind you to keep it right here for more action on the cube your leader in Tech coverage [Music]

2023-06-14 03:44

Other news