This Israeli company wants to optimize your cloud computing buying for better ROI
Ask any DevOps engineer about the challenges of building a large-scale infrastructure that can handle their company’s computing needs, and the issue of managing costs for cloud computing space is sure to be near the top of their list.
In looking to tackle this constant balancing act, Spotinst has developed a pair of products to help companies optimize and monitor their capacity buying for an overall better ROI. Spotinst claims that they can save their clients between 70-90% on their compute instance costs from providers like Amazon Web Services, Google, and Microsoft Azure. On Wednesday, time will tell whether their company will be the best among the ten startups participating in the Geektime Conference‘s Startup Arena pitch competition.
CEO Amiram Shachar, COO Yakir Daniel, and Liran Polak, who serves as their Chief Architect, founded Spotinst in February. Grown out of an award winning research project during their studies, their service launched in July and is now in use by roughly 30 clients.
At this point, Spotinst has two products on offer for clients, Elastigroup and Elastitude.
Getting the most out of the cloud
As companies continue to scale up their operations, they require significantly larger amounts of computing power. This can be for running processes on virtual machines and other needs that are essential to their business. However, building up and maintaining this kind of infrastructure can be costly and in most cases unnecessary since the majority of companies’ needs can be highly dynamic, constantly shifting from month to month. For many companies, it simply makes better sense to rent server space.
Stepping in to offer solutions to meet this need are the giants of the digital age with their massive data centers, who are more than happy to rent out their excess capacity at an hourly rate, of course.
The magic of Spotinst’s algorithm is its ability to reverse engineer the calculation of the capacities held by different cloud computing services like AWS and Google to predict prices.
With Elastigroup, Spotinst’s algorithm is constantly checking across the different service providers to gauge their capacity levels, and then buy that server service for their clients based on a supply and demand model. Like all markets, when there is a lower amount of supply in a data center, the prices go up, and vice versa. By combining the knowledge of where the providers stand with their levels of capacity with price comparisons from each one, they can save their customers some serious cash.
Their second product Elastitude is aimed at tracking computing instance usage and needs based on a defined set of metrics either before or after they start renting capacity. With this product, they can analyze the needs of the client and optimize future buying based on the network, CPU, disk I/O, or any other third party monitoring metric requirements for the workload, seeking out the best instances per case.
Spotinst offers two kinds of pricing models for these services. The first is pay for performance, where they take a 20% cut of the earned savings for the client. Alternatively, they have a subscription service based on a fixed rate for the hours that customers expect to purchase over the course of a quarter or year.
In speaking with CEO Amiram Shachar, he explains the motivation behind the project saying that, “We understand from our previous experience working in the field how important it is to find exactly the right processes to handle our workloads. With such a wide range of instances, it can be difficult to find the right ones that serve different purposes for handling the workloads, and how important it is to choose exactly the right ones for the job.”
How do they stack up?
Regarding competition, a company called ClusterK had a similar product during their two-year run before being acquired by AWS in April for an undisclosed amount between $20-50 million, and are now developing these mechanisms inside of Amazon.
Another one is Batchly that is also working with the capacity pool market, but appears to be focused on big data and offline processing due to the risk. However, they only list that they offer services for AWS, whereas Spotinst crosses multiple providers, giving the win to the newcomers.
Spotinst brings with them a very talented team and a lot of confidence. By promising their clients that they can meet the capacity needs with their algorithm for mission critical applications, they are stepping into a very ambitious position that could pay off big in the long run. Based on the success of the ClusterK exit, the potential payout for this type of venture could be equally lucrative as well. So far, Spotinst has raised a $600,000 seed round and are close to the end of their Series A, which is expected to be in the millions.
Their vision moving forward
Shachar stresses the importance of promoting the cross cloud architecture utilization concept as a new standard in the industry. Buying capacity should not be a sentimental issue he says, explaining that the team’s focus is centered around getting their clients the best bang for the buck from across the different providers.
One major step that they are making in order to reach this vision is the upcoming release of a live migration framework for containerized applications like Docker.