Komprise aims to kill data sprawl with virtualising software

Rubrik, Primary Data, Formation Data Systems, and Hedvig: all startups that want to provide a virtualising data abstraction layer above and across multiple storage silos to make storage access and management easier. Now you can add Komprise to this list.

None have done it, yet. All are led by founding teams with track records and credibility. All are backed by VCs reckoning they have a chance at generating gold nuggests from the virtualising data management seam.

Komprise was founded last year and has just got itself $6m in A-round funding from Canaan Partners. The three founders are CEO Kumar Goswami, president and COO Krishna Subramanian and CTO Michael Peercy. This trio set up Kaviza to get rid of SAN storage in VDI and sold it to Citrix in 2011. They had previously set up Kovair which is alive and prospering today.

That's a good track record.

What Komprise aims to do is provide data-management software, offering a corporate-wide, indexed virtual repository of data in both on-premises and public cloud silos. This software will provide access to the data, as well as managing, organising and protecting it. The management will be automated and driven by analytics.

Existing applications will use existing protocols to access their data and the existing data infrastructure will remain. There are no agents running on separate storage resources or in the accessing applications. What we have here is a control plane to the side of the data path.

There is a focus on unstructured data and Komprise says it will reduce the management and cost burden of buying and operating existing storage by up to 70 per cent. It will eliminate migrations and over-provisioning, and reduce on-premises backup size – solving world hunger will presumably come in version 2.

Commentary

One issue here is that such panoramic, all-in-one data management software tools would seem to compete with the secondary data virtualisers, like Actifio, Delphix and Cohesity. Possibly the Komprise-like startups would say that's just another silo, albeit virtualised, to manage.

I don't think so. It won't be that easy and there's a looming conflict in data management approaches coming our way.

Back to data management layers: we have five startups here, all founded by smart people with excellent credibility, all looking at solving the data sprawl problem with an overarching software layer involving much metadata munching. All are funded by smart VCs out to make a buck - or several million.

Silicon Valley is awash with cash, full of bright-as-a-button people skilled at recognising IT problems and building software tools to solve them. They can't all win.

It's like a frenetic startup bazaar, from which good stuff will soon enough emerge. We just don't know yet which companies will break out. ®