Spotinst Snags $35M In Funding From Highland Capital, Intel

​

Spotinst said Tuesday it had secured another $35 million in funding to extend a technology that helps enterprises save money by tapping an abundant cloud resource—unused compute capacity.

The Israeli-American startup was founded three years ago on a vision of using artificial intelligence to enable customers to better take advantage of virtual machines that only run when the cloud provider has idle capacity, known as spot instances.

With the Series B Round, which brings total funding to $52 million, Spotinst will extend that technology across hyper-scale clouds, as well as containerized and serverless environments, founder and CEO Amiram Shachar told CRN.

Highland Capital Partners led the round, joined by previous investors, including Leaders Fund, Intel Capital, and Vertex Ventures.

[Related: AWS Releases Next-Gen ‘Burstable’ Instances]

Excess compute capacity is this "great resource in the cloud," Shachar said. Providers like Amazon Web Services, Google Cloud Platform and Microsoft Azure "built giant data centers" that often have unsold compute cycles.

The three hyper-scalers have different models for offering machine types that run only when resources are idle—AWS calls them Spot Instances, Google calls them Preemptible VMs, and Microsoft more recently introduced Azure Low-Priority VMs.

While the cost of those cloud servers typically comes out 80 percent lower than comparable instances, the catch is the providers can evict users at any time that capacity is in demand.

Because of that unreliable availability, spot instances are popular for batch computing jobs that can run whenever there's a window, but are rarely used for production workloads, especially those mission-critical.

But Spotinst's Elastigroup platform looks to make spot instances production-ready with a prediction algorithm that "can identify when the cloud is permeating this capacity, predict when it's going to happen and migrate applications from one instance to another."

The company was founded in Tel Aviv, Israel, but has now opened offices in San Francisco, New York and London. A customer roster including Sony, Samsung, Unilever, Qualcomm and Ticketmaster demonstrates the solution has appeal to large enterprises looking to reduce costs.

"This gives you the most value in the shortest of time possible," Shachar said.

The platform's use of predictive algorithms, analytics and historical data predicts when a spot instance is about to be terminated, then moves the workload it's running to another virtual machine, avoiding disruptions.

The software, deployed within the customer's cloud account, understands the underlying architecture of the workload, such as web-services or a database, and migrates all necessary components, like load balancers, Shachar told CRN.

The solution currently is available only for AWS, but the latest funding will allow the company to integrate Google and Microsoft clouds. Product development will also extend into container frameworks and serverless compute platforms, which are fertile new territory for the startup, Shachar told CRN.

In addition to expanding onto new clouds and compute environments, Spotinst will invest in hiring more workers to accelerate sales, and look at some potential acquisitions, Shachar said.

The company already does 30 percent of its business through the channel. Its best partners are boutique consultants and systems integrators deploying AWS for their customers.

But Spotinst is eyeing "the next engine of growth," he said, "which is working with managed services providers."

Bringing into its channel cloud partners that deliver ongoing management of enterprise infrastructure "is going to be the next step in dominating the market of workload management," he said.