10X Cloud Management with Automation

Welcome to The Spend Advantage™ Podcast by Varisource, the Savings For Your Business.   Get access to discounts, rebates, benchmark, savings for renewals and new purchases for 100+ spend categories automatically for your company

We interview amazing people, companies, and solutions, that will help you 10X your bottom line savings and top line growth for your business --- https://www.varisource.com 

Transcript

Welcome to the Spend Advantage Podcast. Spend advantage is the competitive advantage for your spend across 100 plus vendor categories. This podcast is all about interviewing amazing people, company, and solutions that will help you tend your top line growth as well as bottom line savings for your business. Hello everyone! Welcome to another episode of the Spin Advantage podcast with Vera Source. I'm super excited to have one of our new partners on the call today, Dmitri, who is the CEO and founder of Emma. Emma is the all in one multi-cloud management platform to help your business manage and save on all clouds. Welcome to the show, Dimitri. Yeah, thanks. Thanks a lot, Viktor. Thanks for having me. Pleasure being here. Yeah. You guys have such a unique. Well, there's two things. One is we love cloud savings of all kinds, and you guys have a very unique solution. And so excited for our conversation. If you don't mind, maybe give the audience a little bit of the company founding story. Uh, would be it's always fun. Yeah, sure. Happy to. Uh, we've started to build in the product back in 2018, as a product was born out of a real pain. Uh, deploying and managing infrastructure across the clouds was really slow. Uh, error prone and expensive. And, I mean, it's still expensive. And we wanted a better way to orchestrate this infrastructure. We wanted to make it, like, really fast, safe, cost efficient without getting locked into one vendor. So we decided to start building, uh, a product. And, I mean, we've spent nearly. Five years, uh, like four years before the first prototype that we released and literally six years before we we we we realized that we, uh, have hit the, the product market fit. And, uh, the customer finally found the, the right approach to our customers and the customers. They finally realized that if they use AWS, Azure, and Google simultaneously, they they they're actually adopting multiple clouds. And they are not like a single cloud user that has a kind of a extensions in a, in a different other providers. So, I mean, we've done a lot to educate the, the community. We've done a lot to explain that. Guys, uh, there is a real problem and AMA is is a right solution. Yeah. So what what what did you do maybe in your career before that made you, were you always in cloud, um, or what made you want to get into kind of solving this, uh, area of cloud? Yeah. You know, um. It's a great question. So the thing, the thing is that I started my career like, uh, as an engineer, like network and as a security engineer, and I was a bad engineer. So that's. That's why the guy, the the guys I was working with, they told me, look, you're a bad engineer. Uh, you got to do something. Something else. I'm kidding. So I wasn't that. I wasn't the best engineer. However, um, I always liked, um, building the products. And when I faced the opportunity to start building the product at one of the companies I worked for, uh, I asked for transition. And then I was, like, managing the product. And it appeared that that was like a first, uh, networking switch that, uh, is being like, ten times cheaper than than Cisco's. Um, I would say competitive equipment. Uh, so we have built out that switch and we released that. And, um, then I was the only person, uh, who was able to sell this. So this is how I moved from the product creator to the sales guy, and then guys out of, um, out of one of the biggest American networking companies. They. They noticed me and they offered me a job. So I was, uh, leaving the team, uh, responsible for one of the top 15 global accounts for for this company. And. Yeah, that's that's how I became like, I found myself managing the sales teams, and I was quite successful. I was, uh, leading the sales teams in the different countries and in different regions. And in 2018, I realized that there is like a problem in the in the cloud space because I was working for for the cloud networking company. They they were and they are actually building the the networking switches and the switches. They, they support the data center interconnect. And our customers, uh, that days, they were constantly asking us if we can also help them to interconnect their on premises environments, on premises data centers, uh, with their like newly adopted, uh, cloud environments. And the the obvious answer was, I mean, we we cannot help you guys. Unfortunately, we are building the networking equipment and you got to find another solution. And when you hear this really often from the different organizations, really big organizations, uh. The biggest players out there. Um, and if you have this, you know, kind of like soul of an entrepreneur, then you, you're like, okay, it is it seems that there is a problem and maybe I should leave my, um, everyday job and start building something. That was a big mistake, by the way. Um, so, yeah, that's that's how we started the platform. So I decided to leave a company and, uh, start the new venture. And this new venture, uh, became Emma. Yeah. I, um, I mean, obviously as an entrepreneur, like, I just love hearing these stories, you know, obviously, um, of what happened and everybody's journeys a little different. Uh, but one word you mentioned that I feel like, um, is going to be a big topic for you and me today is education, because most companies are going to the cloud, especially, you know, startups or tech companies, everybody starting in the cloud now. So people are familiar with the cloud, but most companies are only using one single cloud. Um, but that usually causes vendor lock in and other limitations, but they're not aware of it. So the word you mentioned is education is can you kind of maybe describe why are most companies using single cloud today, uh, and not multi-cloud? Uh. I would say that, first of all, if we look into the the statistics, we'll see that, uh, in the United States of America, 95% of a bigger organizations, they are hybrid cloud. So they are not even single cloud. They are hybrid cloud, meaning they have an on premises environments. And they also, uh, have cloud environments, private private cloud. Public cloud environments are in use. Uh, so that's, that's that's the great starting point. Uh, when we're talking about the, the private clouds, I mean, it's obvious you have your data centers, right? And, uh, I mean, especially with this for these large enterprises, they've been hardly invested into this. But when it comes to, to to public environments, it is also, I mean, interesting. So normally companies, when they start with the cloud adoption, they start with the, with something that they are interested in. So for example, they want to roll out their services faster. And the cloud seems to be a nice, um, supportive environment. Right. So and then they think, okay, we gotta hire someone and they hire someone and they, someone has a previous experience with, I don't know, Microsoft. And then someone spins up the infrastructure in Microsoft and they stay in a single cloud and uh, they like I mean, they're moving more and more of their workloads into the cloud and they rely on a single provider. It creates dependencies. And then they found themselves limited by the pricing tool in regions, availability zones and the roadmap of this provider. And. They realize they. At one point, this provider decides to increase the prices by 10%, and you realize that your cloud environment worth 500 millions of United States dollars annually and 10% increase is like 50 million for you on top. And you're like, okay, we're locked in. But you were happy before you liked the managed services. You never cared about the downtimes. Uh, you never cared about the the fancy editions that other providers, they, they introduce. You were happy, but again. So one day something happens and you realize that you were you are locked in and you cannot do anything with that. And why I'm saying that 95% of organizations, they are hybrid cloud because it is the beginning of being like multi-cloud, because multi-cloud is more than one cloud in use, right? And normally what people do, they move the workloads. They start moving the workloads from on prem into the cloud. Then they realize that the clouds, they are not that cheap. They have their limitations and they start either moving the workloads into another provider, the cheaper one, or they try to, you know, like redistribute the workloads, put in their at a no test and environments at the cheapest provider. Production environments are the most reliable provider. Uh, they also want to keep sensitive data stored on premises. And then they realize, okay, it's a way too complicated. We gotta repatriate everything back on prem. And I mean, that's hilarious when you see all, you know, this, this flows back and forth into the cloud, out of the cloud, between the clouds. So getting back to to your original question, people, they choose single cloud because at the beginning they love all this and all these services. They love the flexibility. They love that there is someone bigger who have created, who has created this, this multiple managed services where you don't care. You just simply ship your code into somewhere. And this somewhere does all the magic. Yeah. I mean, I love how in-depth, like, obviously I love your expertise in these areas. So we can we can talk for hours. Um, you know, and like you said, that's why AWS a lot of these cloud guys, you know, start with the startups, right? So it's like, oh, you're just starting a company. So you're let's get you into the cloud and you're like, wow, I can just focus on building my software. And I don't have to worry about infrastructure and they take care of it. But as you get bigger, uh, like you said, it just becomes so hard and so dependent on each cloud. Um, but I guess educate me because our audience, um, have, you know, startups, there's also midsize companies, and then there's also enterprises. And so I want to make sure we kind of explain to each of these audience. So I mean, what are what do you think are some of the main challenges of using multi-cloud today without a solution like Emma, uh, you talked about, you know, management, right. And complexity. But can you give some examples targeting kind of different audience meaning if you're a startup, what are some of the challenges of using multi-cloud? If you're a midsize company, what are some challenges? First of all, let's be honest. Nobody wants, um, to manage multiple clouds. Nobody literally wants to do this because it's a way too complicated. There's different providers. They have different services. There are a lot of, um, different complications, different APIs, different tools, different resource types, inconsistent policies, hard to track spend usage. Uh, different, I don't know, billing APIs, protocols, different billing cycles. Uh, every cloud works completely different with a different virtualization hypervisor and even with a different hardware equipment. So nobody wants to manage this. I mean, period. And we had a struggle at the beginning when we were pitching the multi-cloud, because when you're reaching out to bigger organizations like bigger enterprises, Walt Disney, United States Postal Services, they tell you, look, guys, I mean, we are multi-cloud, uh, because we have a lot of regions, uh, where we we are present. And, um, I mean, we are building the reliable applications. However, we don't really want you to pitch us the benefits of being multi-cloud. We hate being multi-cloud. It's a way too complicated. And these guys, they have hundreds and maybe thousands of engineers. So but what everyone really wants. Everyone wants to control the vendor. Everyone wants to be cloud agnostic. Everyone wants to decide which services, tools. Uh, applications. Uh, they. They want to use or they need to use. And this is this is important difference. Being cloud agnostic meaning independent is a way different of being multi-cloud. Uh, and that's what what we try to promote and that's what we, we try to explain, uh, take the startups as, as an example. So the startups they normally kick off their, their venture, uh, and they're like building the products, uh, picking up simply the, the provider because of the amount of cloud credits they receive. And they become multi-cloud when they start moving their workloads between the providers collecting these cloud credits, at the end of the day, they have a distributed infrastructure, quite often managing different environments. I mean, they they raise the funds, they hire DevOps engineers, these DevOps engineers, they are building, uh, services. These services are distributed between the providers. And but when they burn all these cloud credits, they are venture firms. They need to cover the real cost of the cloud. So they need to pay for this infrastructure. And then, I mean, everyone is in trouble because, uh, we had these statistics from one of our venture partners, uh, one of their startup startups. They spend nearly 15% of of the round ticket covering the, the, the infrastructure costs at at the different cloud service providers because they simply collected all the available cloud credits. And they, they I mean, they found themselves locked in managed services of these providers. So they've been using like, you know, Kubernetes from Azure, managed Database from, uh, AWS alongside the, um, the CDA and etc., etc.. So, uh, and when they realized that their, I mean, their venture firm and uh, their board members, they were shocked. Imagine you, you just you just raised the funds and you and 20% just vanished. Uh, so that's that's, first of all, why it's, uh, why nobody wants to to intentionally introduce multi-cloud. Uh, they everyone wants to be, I mean, cloud agnostic with the startups, uh, it is quite easy to, to become multi-cloud, not on purpose, but because of this, uh, cloud credits. I would say hunt. Uh, on the other hand, it's really, really hard to, uh, to get rid of those vendor lock ins. And I can elaborate even more on on the topic why enterprises or bigger organizations, they become multi-cloud or dependent on the multiple cloud service providers on, on, on with the examples of from from a bigger players out there. Yeah, I, uh. I mean, what you just said right there. I, um, I wish I could. I mean, it's such a great insight into private equity backed companies, um, and how, you know, again, it's like these vendors are just so good at creating business models to not necessarily trick the companies, but it's like they know. Right. They get you in there and then and then you're kind of stuck with them. Once you use up all the credits and you can't turn it off because your company depends on it. And so it's I just love the way you kind of describe that. So we've talked about the challenges. Can you give maybe three of the benefits of using multi-cloud that companies may not be aware of I guess. Okay. So you decided to to to narrow it down because I'm using too many words when I'm trying to answer your questions. No, no, no, these are great insights. Yeah, yeah. I mean, uh, top three. So first of all, the benefits of being, um, cloud agnostic or, or being a multiple cloud user, uh, cloud user. So first of all, it is, of course redundancy and resilience. So you don't have a single point of failure. Uh, even with our platform, uh, we distribute our workloads across the cloud. So your data is, uh, securely stored at the different cloud service providers, even if EMR. Fails. I mean, if if there is an outage, which is impossible because Emma is a set of microservices. Your data is securely stored at the providers of your choice. So I mean, we can commit on this, uh, redundancy and resilience, uh, of your applications. Second one, of course, it is a cost optimization. Right. So you can use and interconnect the best available instances, application managed services from the different cloud service providers for each and every workload that you have in your environment. Uh, you can simply scale your clusters, picking up the best available, I don't know, spot instance on the marketplace. And you're you are not limited with a single provider or availability zone or availability of these instances. So you you you can choose from any, any of them. And the third one, uh, I would say it's a combination. Right. So, uh, you can deploy in a specific regions or in specific types of the instances if you need to be compliant to a specific regulations, for example, I mean, in Europe, uh, if you're a financial institution, you should be compliant to certain regulations. And the United States of America, you have this, uh, HIPAA certification for the medical organizations, right? Where where you need to use a specific instances. Uh, but they are quite expensive and not, uh, every single application or the part of your environment requires, uh, this HIPAA certified instances. So you can create the mix and, uh, you can be flexible. That's why I'm saying that this flexibility and compliance, they go together uh, with in the euro like choices. Right. And this automatically also improves your, uh, your cloud bill reduces your cloud bill so that that's. These are the these three major benefits. So redundancy and resilience cost optimization and flexibility. So I want to I mean obviously our podcast is called Spend Advantage because um we feel like a lot of times most companies spend money on things and they don't fully know how to optimize it. And um, so I want to deep dive a little bit into the cost optimization part of it. You kind of gave some examples, but obviously let's, let's say I have AWS and there are tools or, you know, resources from AWS that supposedly help me do costs with AWS. Um, but I don't think most companies think about or know that, hey, you can also optimize your cost of cloud by diversifying it. So how can you give maybe some examples of how Emma can help companies optimize their cloud spend? Yeah, of course, for sure. So, uh, an AWS uh, is a great example because, um, I'll give you like a real life scenario, like real life use case. Uh, imagine there's, there's, there's a customer of ours. And these guys, they, they're building electric vehicles, uh, like Tesla, but it's not Tesla. So they're building like, electric vehicles. They they do this, you know, design, uh, they try to understand, uh, how the aerodynamics effects, uh, is affected, uh, with, with the diameter of the wheel or, or like bodykit, specific bodykit, how the range changes, uh, if you use the different types of type of wheels, etc., etc.. So they normally do this, these calculations and they have reserved instances at AWS. They were 100% AWS, uh, reserved instances at AWS. They they were using preemptive instances at AWS. They had commitment plans. ET cetera, et cetera, etc.. And the entire environment, uh, has been built there. But what they also noticed that, for example, there are different cloud service providers, even that even in the same regions or like nearby regions where the the compute costs lower than the one that they can afford with all these commitment plans, uh, reserved instances, etc.. So but I mean, they they never. Had an idea how they can expand into this, into this class, because for them, it sounded like you got to spin up the whole new setup for these environments. Then you need somehow simultaneously track the availability of these instances, and then spin up the specific environment to be able to use these cheaper instances. And for them, it's like, you know, you pay double. Uh, so what we've done for them, we put their workloads inside the microservice, inside the Kubernetes cluster. And this Kubernetes cluster had actually has an autoscaling policy that says, in case you need to get more compute, you scale your worker nodes on the instances like the cheapest available instances on the marketplace, no matter what cloud service provider, uh, offers you this, this, uh, these instances. And, uh, you pick the just the spot instances like 100% of the spot instances, no matter if it's Azure, AWS, Google, Oracle or whomever else offers this, and you scale the workloads. And when you don't need this, this, uh, instances anymore, you simply release them. So we've managed to save 87% for these guys out of their cloud bill. Just simply moving their into these guys, moving them into the Kubernetes that scales across the cloud service providers, and simply picking up the cheapest available instances and releases this these instances where when they do not need them anymore period. That more of a consulting or professional service from you guys or is that something they use the platform and it kind of automates for them. No, I mean, our platform is fully automated. You can spin up the Kubernetes cluster, you can predefine and automate this, uh, this autoscaling policies. So your cluster scales when it needs to scale, it scales up, scales down. You can create different schedules, uh, when you shut it down, when you spin it up so you can automate everything and you basically spend minutes, not even hours, uh, spinning up these environments and make these environments scalable. So here's a here's an interesting question. I come across a lot. I want to get your thoughts. So obviously, uh, do you feel like Emma can how can Emma best help companies and kind of give you two scenarios? One is company that doesn't have a lot of cloud engineers and expertise. And maybe maybe they don't know the clusters and everything. The best way to optimize. So maybe they have one cloud guy, maybe is the developer and the cloud guy. And then you have obviously larger companies where they have more, you know, I guess, uh, expertise or experts on cloud and they're optimizing it, doing FinOps and everything. How can Emma, is Emma better for, uh, the, you know, for can I cover both use cases or is it better for certain? Uh, okay. Can you kind of talk through that? We can cover both use cases. So, uh, I mean, take as an example. I really like the, the real life examples. So, uh, we have a number of customers. Uh, they are like a software as a service companies that have grown beyond a simple cloud setup but don't yet have a mature platform team. So their DevOps engineer, uh, the DevOps team is small and overwhelmed, and infrastructure is growing in complexity. So normally what happens, uh, and the the major pain points is a lack of automation, usable configs, fragmentation, tooling, etc.. So what we can do, we can act, uh, as, as the platform that automates your cloud operations. So your like DevOps, like engineering team, this smaller team, uh, like really small teeny team, uh, can focus on a product, not on the pipelines. Right. And we can take care of this infrastructure. We can take care of the deployment, automation, provisioning, uh, extrapolation. We also give you, uh, recommendations how to to reduce your cloud bill, how to improve your infrastructure based on the behavior of the workloads. So you, your smaller team can focus on something else and forget about this routine, dealing with the the infrastructure management and deployment and automation. So that's what we do for for the smaller organizations. Uh, when we're talking about the bigger organizations where, uh, people they want to, to, to to make, uh, decisions based on the analytics, on the data. Uh, so, like, I don't know, see, force and the financial decision makers. So first of all, we, uh, we know that they want a transparency. We know that they want a predictability and efficiency in cloud usage. And they care about this because for them, it's like, you know, it's a black box. They do not understand what's going on there. And normally we know that developers we have it in our in our company developers and engineers. They simply have this, uh, this I don't know, like corporate credit cards. And they pay for whatever cloud resource they want. And then our finance team is overwhelmed with this invoices and they do not understand why do we need, like, hundreds of the VMs? Uh, is it on purpose or what? Or we forgot about these VMs. So, uh, the pain points here are the cloud based overspend. And because of the misconfiguration or idle resources, uh, and the growing DevOps costs. Uh, so, I mean, for them, these are the pain points. And we can cut this cloud waste and we can prevent the misconfiguration, we can track these idle resources. We can replace complex tooling with a unified orchestration. Orchestration level. Orchestration level. Uh, where layer where? Um, we are saving not only the time, but also the money and the DevOps headcount. So you can actually, um, start making decisions based on the real data that you, you have in front of you across the variety of cloud service providers and in the, in the understandable UI. Yeah, I first of all, I love the way you give examples and talk through this because I can see where your technical you're an engineer so you understand the the product side of it, the technical side. But then you're also a good salesperson. And so you're able to tell stories, you're able to explain, uh, in a sales way. So it's, it's a, it's a great combination. And uh, yeah. No, I think, you know, everything you mentioned makes a lot of sense, uh, as we kind of come to the end of, um, close to the end of the conversation here, there's a couple last questions. So we talk a lot about, you know, optimization. Uh, let's let's go talk a few seconds about government governance, governance and compliance of cloud. Why is it better and why is it better, uh, with Emma versus the cloud platform using the cloud platform themselves? You think? Oh, it's a great question. So cloud platforms, they have their cloud native tools or I don't know, services. And these tools, they only govern their own ecosystem. Uh, what is important to understand that AMA gives you the centralized decentralized policy enforcement across all of these environments. All these services and all of these clouds. Uh, we have an embedded tools for the continuous compliance monitoring. You can decide yourself where you want to store the data. If you want to have this data accessible, uh, to the user. Uh, I don't know, providers, instances, uh, people, I mean, team members environments. So it's up to you. You can physically because we have a physical network and backbone that interconnects the cloud. You can physically To disconnect one provider from another or one region from another. You can enable or disable um services in a specific geographies. Uh, our GitOps native workflows, uh, they they they are natively ready for any sort of, like, audits. And, uh, our platform is like multi-cloud. I would say it's multi-cloud, native by design with a self governance and and a compliance models. So that's that's why it makes us, I would say different to a standalone cloud platforms. Right. Uh, so the last question, obviously we can't have any conversation without AI. You know, the AI in the last year and a half, two years, it's been crazy, the pace and everything. So as we kind of wrap up the conversation here, what is your thought on AI in general, I guess. And then two is how, uh, you know, where's AI may be impacting Emma? Um, from a business perspective. Um, so. Yeah, I. Is, is is a great topic. And, um, I mean, the tools they have dramatically changed, uh, over the last couple of years. Uh, when, when OpenAI introduced the first version of a ChatGPT, I was I was a huge skeptic. Now, I mean, the things they have evolved and, um, what we can do and what we do, uh, we can interconnect the different unique, um, AI services that providers they offer. So, for example, AWS offers for the compute or GCP offers for, for AI, Azure, for for the enterprises. So you can mix and match in a single interface all these different, uh, services for the different purposes from the different worlds without managing the chaos behind the scenes. Uh, on top of that, we can and what we allow to, to, to build with our platform is a logical GPU clusters to, uh, not to train the model, but for inference purposes, because to train your model, um, you need to be sure that the latency is less than a couple of milliseconds, which is impossible if you are connecting the clouds. Only the cloud service providers are like literally have their equipment on the same building. However, for the inference purposes, we can create the logical clusters where you can scale your workloads, picking up, uh, the GPU instances across the different cloud service providers. So you can fulfill literally the entire need of your company with the accelerators, uh, from the different providers. And even, uh, you can move the workloads from one provider into another in case your, our platform, uh, finds something that is like significantly cheaper. Uh, if the, the resource if, for example, you were using the spot instance with the GPU accelerator and if the resource is taken with the attached the volume, we do the snapshot of your memory, and then we spin up another instance and we put this back into the new instance so you don't lose, uh, the results of, of your, uh, of your processes that that's been running there. Um, that's what we did on top of that, I mean. We we can interconnect also your on premises environment with the cloud environments. For example, you uh, you leverage your servers with the accelerators. Uh, you I don't know, you do the training and then for inference purposes, you need more compute capacity. So you can roll out the workloads into the cloud through our platform with your, um, private environment is, um, is is connected to, uh, to our networking backbone. So we do a lot of things for the AI company, starting from the provisioning and ending up with the, uh, the services that are scalable and across the variety of cloud service providers. And we also can interconnect these services with the native services that providers they offer you. Yeah. I, um, I mean, there's so much value that you guys provide. That's why we're so excited to partner with you and and help you grow the US market. Um, so the last question we always asked every guest, uh, that comes on is you, you've seen a lot. You've done a lot. Um, if you have to give one advice, personal advice or business advice that really passionate about, uh, what do you think that would be, Dimitri? I think it's. I mean, for, for entrepreneurs, it's the obvious one. Like, don't give up. If you really believe, uh, you're building something that changes the world, don't give up and simply do this. Build this. Uh, no matter what happens, uh, around. So you you gotta make it happen. You're an entrepreneur. And don't listen to to. I mean, to anyone that there are only customers. Uh, and quite often the customers they run, uh, but on the other hand, there are only customers who can give you, like, real feedback. You you gotta not even listen to this feedback, but you gotta take this feedback into consideration and build your product or shape your product, uh, with regards to your to your vision. And and at the end of the day, you'll get something that, uh, that is really incredible, that scales, that has an army of, I don't know, followers, users, etc. so never give up. And you, you, you will achieve all the heights that that you want to achieve. I I love it. Man you are. Uh, no. This has been a great conversation. We really appreciate the partnership. And, uh, thank you for being on the show. Yeah, man. Thanks for the invitation and super excited and partnering with you guys. That was an amazing episode of the Spend Advantage Podcast, where we show you how we can help you, techniques, your bottom line savings, and top line growth for your business. Hope you enjoyed the conversation, and if you want to get the best deals from the guest today, make sure to send us a message at Sales at @varisource.com