Cohesity has launched an application marketplace for its secondary storage systems that provides a way to run third-party apps on the storage devices themselves, much as you run apps on your smartphone. It’s a really clever move.
“Data protection is just an app,” said Mohit Aron, Cohesity Founder and CEO. The intent is for Cohesity to be the place where your data lives, and you then bring applications to the data, instead of moving data to the apps.
We’ve previously seen efforts from all kinds of vendors to ‘unlock’ the value of data sitting on their devices, but it’s especially common with backup vendors. Backups are frequently seen as a kind of dead weight, a cost with no benefit (though the benefit quickly becomes apparent when you need to restore data) so vendors have long sought another way to provide relevance lest they continue to be treated as insurance sellers. “We already have a copy of the data,” the thinking goes, “so why not do something else with it?” Yet that something else never turns out to be all that compelling.
I think Cohesity’s approach may finally lead to something that is.
Rather than rely on your single backup vendor also having the spark of inspiration that finds the killer use-case that is relevant for their entire install-base, Cohesity doesn’t have to. They let other people figure that out, and just as some people love to play games on their phones while others can’t get enough of social media apps, Cohesity customers can pick and choose their favourite combination of data apps to suit their own data needs. Cohesity doesn’t need to write them, it just needs to provide the platform that makes it easy for application developers to find customers.
At launch, there are some example applications by both Cohesity and certain partners, such as Splunk, SentinelOne, and Imanis Data. The ability to scan data and report on what kinds of files you have, where they are, how much data is consumed by different types of files, these are all clearly useful things to know. It reminds me a little of the vision of DataGravity from some years ago, but with far greater scope.
Because why limit your app to reporting? Why not scan files for malware with the ClamAV app? Why not point SentinalOne’s AI engine at a huge mass of secondary data to train its models?
Why not go further? Why not train other machine learning models on this lovely big pool of data you have sitting there already, handily close to the compute needed to run it? Why go through the hassle of trying to move massive datasets to compute in the cloud when you can bring compute to the data and run your app there? It’s quite compelling.
In my opinion, Cohesity has undergone something of a resurgence in the past year. Early on, the company struggled to get as much attention as some of its equally well funded rivals in data protection, but its technology attracted plenty of customers. Now it seems the marketing is starting to catch up to the engineering and the potential here is genuinely huge.
There is already massive spend in the enterprise on data analysis tools and techniques, and with the hype around AI and machine learning that’s not changing any time soon. Cohesity can do very well just by being able to build an ecosystem of software providers keen to make it easy for customers to use their software on their own data. If other developers can easily create apps, list them on the marketplace and get paid for doing so, then I can easily see a solid market developing just the way the Apple App Store did.
I can’t predict what the big apps will be, any more than others could predict the popularity of Flappy Bird, but I feel confident in suggesting that there will be at least one, and probably several.