Traverse City Record-Eagle

Michigan

Related Photos

  • DeepField founders Joe Eggleston, left, and Craig Ladbovitz pose at the company's office located in the Tech Brewery in Ann Arbor. The Associated Press/Jeffrey M Smith

    CSS2C8F.jpg
September 16, 2012

Michigan-based tech company launched

Techs found firm based in Ann Arbor

----

ANN ARBOR (AP) — Some companies provide tangible products or services that are easy to understand. And then there are companies like DeepField. The recently launched Ann Arbor-based startup that analyzes big data from the cloud to help content providers and carriers make smarter business decisions.

"We've developed something we call our cloud genome technology, which is pretty groundbreaking," Naim Falandino, a recent University of Michigan masters graduate and DeepField's chief data scientist, said.

DeepField already has a base of paying customers and private beta users. Falandino said he could not reveal companies currently using the system, but they include major internet companies. Data gathered by the company's technology has been featured in Wired Magazine, PCMag.com, and Computerworld.

The company is also venture capital funded and has investors that range from a fund in Houston, Tex. to RPM Ventures here in Ann Arbor.

Tackling the daunting challenge of producing a system that could analyze massive amounts of data produced by servers in all corners of the globe was the brainchild of two men who were also on the ground floor of developing Arbor Networks.

Craig Labovitz and Joe Eggleston had both been at Arbor Networks for approximately 10 years in spring 2011 when they left the company to found DeepField.

"The real problem we identified is that everything is changing in the way companies build stuff on the internet. (The internet and the cloud) are the most complicated thing we've ever built in human history," Labovitz, CEO of DeepField, said.

"Understanding how the money works, understanding how the pieces fit together is just very hard today."

While 10 years ago, a website would be housed on an in-house server, today pieces of different web operations come from all over the world.

"A while ago AnnArbor.com would have had a closet where a server that literally held AnnArbor.com would have been stored," Labovitz said.

"The way you build stuff today you have a piece from Amazon over here, and widgits from another place over here, and that might be running from a data center in D.C. It's a very different world how these services are built, and you just need one part of the chain to fail for bad things to happen."

Falandino said the company analyzes "flow" data from servers that allows them to essentially "map" the internet and see how different types of data get from point A to point B.

"All services have a traffic signature that functions as its DNA," he said.

The cloud genome technology can learn identify these signatures and then modules that the company has developed help users to interpret the data and make it functional for them.

Falandino said users of the product could range from large companies that want to see how their internal "cloud" is functioning to local internet carriers that want to know how much the transmission of certain types of information is costing them.

"With our analytics we can go down to the user level and tell you how many people are watching Netflix on a PlayStation 3 right at this moment," he said.

Text Only