Cloud analytics startup DeepField Networks has emerged from stealth mode, setting out its stall to offer large organisations a way of tracking and managing the performance and structure of their cloud infrastructure.
The market the company has set itself up to serve is so new even explaining precisely what the VC-funded, Michigan-based firm does can sound unfamilar.
DeepField’s central contention is that every underlying service in the cloud can be fingerprinted using what it calls a “cloud genome”, which can be used to unlock cloud telemetry on what is really going on under what otherwise appears to be an inscrutable wall of traffic and interactions.
Running on virtual appliances claimed to scale well, the service will automatically “discovers” these elements, updating as such services evolve. The end result should be useful cloud analytics that can understand the infrastructure according to service, geographic, and economic models.
The service is high level which aims it squarely at entities with an interest in understanding cloud infrastructure in this way, largely content service providers, carriers but also some corporates with complex clouds worthy of examination.
“The cloud is the largest and most complex technology ever created,” said DeepField co-founder and president, Dr. Craig Labovitz, formely of security firm Arbor Networks, also headquartered in Ann Arbor.
“Services now run everywhere and anywhere, presenting a bewildering array of business, IT, and security options and challenges. DeepField provides the tools for enterprises and providers to navigate this increasingly complex, competitive, and highly dynamic cloud environment.”
The system’s analytical engine was able to work out that up to an extraordinary one third of Internet users come into contact with Amazon’s Web Services cloud at least once every day, to pick one example.
Despite the newness of all this, DeepField is superficially not entirely removed in principle from Labovitz’s last company, Arbor Networks, which applied analytical principles to Internet traffic to block the security threats (such as DDoS) hidden within them. The end purpose remains very different, however and Labovitz is keen to stress that the companies do not compete in any way.
The company had started running with unnamed customers, including a number of cloud operators, with public betas due to begin in August, it said.