Managing a complex multi-cloud environment is no easy task. Emma’s multi-cloud platform wants to make clouds and the capabilities they offer as natural and accessible as electricity. We spoke with the company’s Dmitry Panenkov about this ambition.
Dmitry is founder and CEO of emma (enterprise multi-cloud management application). He sees a great desire for simplifying cloud deployment. Especially when it comes to deploying resources on multiple clouds within the same organization, it is still very complicated. This not only has implications for setting up and managing such a distributed environment. It also causes an organization to have little insight into the financial impact of the cloud. Especially now, with the advent of Kubernetes and microservices, it is all becoming increasingly ephemeral.
In order to illustrate what the market needs, and what emma aims to provide, Dmitry cites the example of Netscape: “Before Netscape, everyone was coding to get on the Internet. When Netscape introduced the browser, everyone could use it.” This same development is going to happen in the cloud. Its capabilities are currently only accessible to a select group of professionals, is his reasoning. That has to change, and emma wants to play an important role in that.
Abstracting away the cloud
So to make the cloud (and multi-cloud) available to everyone, emma needs to build the equivalent of a browser. That’s where Dmitry started about six years ago. After about 3.5 years of building the product, based on many conversations with the market, he introduced emma to the world at WebSummit in 2021. This was followed by a very rapid adoption. Without any marketing, the company went from a handful to 40 employees in a year.
In early 2023, emma did a funding round and raised six million dollars. With this, it obviously wants to continue building its own platform, including hiring more people. In addition, the amount of the funding round also indicates that emma is approaching this in a sensible way. Not huge investments, but investments that the company can pay back. That may not provide the familiar ‘rocketship’ experience you regularly see in mostly American startups. However, it does guarantee that the team can build the platform in a careful and good way.
What is emma?
At the end of the day, emma aims to provide a management platform for multi-cloud environments that ensures no one has to worry about the routine chores, such as setting up and maintaining VMs, clusters, applications, backups and network. Employees can then focus on actually building an application for their customers, for example. The same goes for Kubernetes, Dmitry points out. “If you want to build an application in Kubernetes, all you have to do within our platform is specify where you want to run it, we do the rest,” he indicates. For example, if you like to put an application’s data in MongoDB, you select it in emma’s environment and link it to the appropriate Kubernetes cluster.
The above reminds us quite a bit of what we heard from the CEO of Pulumi earlier this year. That company is specifically about Infrastructure-as-Code, but the concept is broadly the same. The platform provides connectors to all major environments. End users only have to specify what they want to see as the end result, put succinctly. It should come as no surprise, then, that Dmitry is a fan of the approach of Pulumi, which he believes has the edge over a more established player like HashiCorp.
In addition to a “Netscape approach” to cloud infrastructure, cost efficiency also plays an important role within the emma platform. Especially when it comes to Kubernetes, containers and microservices, there are pods and nodes that not infrequently exist for very short periods of time. This offers a lot of opportunities in terms of cost. After all, you can determine where the Kubernetes clusters can run most efficiently each time you spin them up. These so-called spot instances combined with Kubernetes is what Dmitry calls an “ideal combination for emma,” and the savings here can really be very substantial. He cites the example of a supplier of electric cars. “They were able to save 77 percent in expenses,” he states.
The emma platform, however, is more than a smart and cost-effective way to help cloud architects build multi-cloud environments. It is not only a layer on top of the environment, it also ensures that the clouds are connected to each other. This is the result of one of the insights Dmitry gained during his time at Arista. “When we had data centers connected, we were almost always asked if there was also software to manage the hybrid environments,” he points out. It didn’t really matter what the specific architecture was. Organizations just wanted insight into this.
According to Dmitry, the emma platform can provide the management that organizations are looking for. Not only that, it connects environments through its own networking backbone. This component is very important within emma’s vision. Deploying services across multiple clouds may sound good, but if that means getting a lot of data from cloud A to cloud B, it’s not affordable. For that, the fees that cloud providers charge to get data out of their cloud environment are usually much too high. Through a private backbone, customers do not have these costs.
Dmitry illustrates the added value of the private backbone by pointing to one of emma’s customers. This is the “largest research company in the world.” This organization uses Google BigQuery at the same time as ML services in AWS. To do that, it has to send a lot of data from one cloud to another. It does that through emma’s backbone. What is striking in this example is that this customer uses only this part of the emma platform, nothing else. In other words, it is perfectly possible to purchase emma’s offering in a modular fashion.
Two different use cases
Dmitry and his team are building a platform that could be of interest to many organizations. It has to be, because a multi-cloud strategy is basically an option for all possible organizations. As such, he sees two general use cases for the emma platform among existing customers.
On the one hand, there are the organizations with a lot of legacy in a private cloud (on-prem), moving to one or more public clouds. These need to have a solution to manage the complex infrastructure that they create by doing that. On the other hand, there are the organizations that have emerged in the cloud. Those understand how containers work and they are often well-versed in what emma has to offer.
In time, however, these latter companies will also recognize that they cannot do everything in the public cloud, Dmitry expects. This is inevitable as both applications and data get gravity. Think of things that require extremely low latency, as well as workloads that must be handled at the edge, because otherwise far too much data has to be sent to the cloud. That not only causes delays, but also very high costs. “There is a good reason why companies like Snowflake and Datadog are building their own data centers,” Dmitry illustrates his point.
Broad vision for efficient multi-cloud management
Ultimately, emma’s vision is a pretty broad one. It involves FinOps, DevOps and CloudOps. In other words, the platform must add value at multiple layers within organizations. So during our conversation with Dmitry, associations to conversations we’ve had before come up frequently. From Spot by NetApp or CAST AI for FinOps via Pulumi to Symworld Cloud and Portworx for DevOps and CloudOps, we recognize (parts of) those propositions in what Dmitry tells us. It sounds to us that emma wants to add another layer of abstraction on top of that, to make it even more widely applicable.
Right now, we think the FinOps side of emma’s approach is the most interesting part for organizations. If you set up a distributed multi-cloud environment optimally, you can save a lot of costs. The fact that emma also wants to make this very simple is certainly not insignificant. But that, too, has a lot to do with efficiency. Not so much directly on the budget, but indirectly by being able to deploy people more efficiently. In itself, this is a logical approach as far as we are concerned. The cloud can make many promises in terms of possibilities, but if it is too expensive, this will be an obstacle for many organizations to actually use it properly.
emma is a multi-cloud management platform to keep an eye on in the coming years. That is something we can confidently predict. The platform is certainly not finished yet; in fact, it never is. Conceptually, emma is on the right track, we think. Ultimately, cloud resources have to become a commodity. For that, you simply need a platform like emma’s, which combines FinOps, DevOps and CloudOps in a single environment.