Chris Lynch wants to be Big Data's Big Daddy, with plans to back 20 young companies on the East Coast that deal with data sets characterized by "volume, velocity and variety."
Most intriguing among those startups is one that the recently departed head of HP's Vertica Big Data business hints will be launched during the next 90 days and will be of great interest to service providers, among others.
"I'm going to build the high-speed railway for Big Data," says Lynch (pictured), who as CEO helped make analytic database management software company Vertica worth some $340 million to HP last year. "All these companies I back, and others, will run on that railway system."
The "skinny pipes" in the cloud are currently preventing organizations from fully exploiting Big Data systems, says the new angel investor.
Lynch warns that AT&T and Verizon "aren't going to survive very long selling iPhones," so had better learn how to monetize the bits that run over their networks. "I'm going to help them do that. They don't know it yet, but they will in the next 90 days," he says. "I'm going to participate in the economics of showing them how to make a living on those networks in the Big Data world. If they don't, they're going to be relegated to the transport and everyone else on the endpoints is going to make the money."
Lynch is relatively new to the database and Big Data world, even referring to himself as a "network guy." His career path -- initially in sales -- since the late 1980s has traveled from Digital Equipment to router maker Wellfleet Communications to Gigabit Ethernet switch maker Prominet to Web switch maker ArrowPoint to Cisco (which bought ArrowPoint) to file virtualization company Acopia (acquired by F5 Networks) to Vertica and HP. Lynch is passionate about Big Data, but also feels strongly that networking will play a huge role in Big Data's development.
The executive, who credits early punk bands such as the Sex Pistols and Ramones for inspiring him to march to his own beat, says the key thing to know about Big Data is that it is about how organizations can exploit not just the information inside their walls, but reconcile that with structured and unstructured data from outside. "The Big Data problem is less today a storage or compute problem, but rather a transport problem," Lynch says. "People spend so much money with EMC just to store stuff. But how much would you pay monthly to store all the crap in your attic that you never look at? That's in effect what enterprises have been doing and need to stop doing."
He cites Vertica customers such as Groupon and "FarmVille" maker Zynga as shining examples of companies that have gained strategic advantage by making Big Data actionable to stay on top of customer trends in real time. Next up will be seeing more traditional retailers, financial institutions, healthcare companies and others doing the same, and making use of Big Data to address everything from patient care to fraud detection, he says.
"You're going to start seeing businesses leverage Big Data to disrupt current business models," he says. "Over the next three years every single data set, with the possible exception of traditional OLTP apps, is going to have Big Data attributes."
The rise of Big Data is also a great opportunity for IT professionals within enterprise data centers who have visibility into networked applications and end users, Lynch says.
"Businesses that embrace Big Data are going to understand that one of the best sources of data is all the log data in the network," Lynch says. "If I were a network analyst I would jump in and help define what it means to be a data scientist and try to be one for my company. I'd figure out how to stuff that log data into Hadoop, Vertica or whatever platform you have and use tools like Tableau to provide insight into the business through the network traffic. ... Be bold enough to say, 'We ought to connect this data on top of a feed from Facebook, map them together, see what we learn about our business and take action on it.'"
Asked about whether big network companies have a clue about Big Data, Lynch says yes and no. Vertica actually was in talks with Cisco before the HP buyout about OEMing its technology for use on Cisco switches, to provide analytics on those devices to help them make better decisions and capture log data. But Lynch fears Cisco might be looking at Big Data more to sell more infrastructure rather than to make smarter network gear. Still, Cisco's incubation of startups like SAN switch maker Andiamo and stealthy software-defined network outfit Insiemi offers hope that Cisco is thinking more broadly about Big Data.
As for his own approach to starting up companies, Lynch last year gave some tough love with graduates of Bentley College in Waltham, Mass., during a commencement speech at the school where he earned his MBA (see video below). He urged students to take smart risks. The executive says a big reason for putting his money and efforts behind seeding Big Data companies is to help the East Coast regain its entrepreneurial mojo. Lynch says he wants this to be part of his legacy.
Vertica was built in large part through the brains of MIT computer scientist Samuel Madden's disciples, and Lynch says more such students (also from local schools such as Yale University and Brandeis University), could provide the foundation for hot new companies on the East Coast. "I'm at the point of my life, approaching 50, where I feel my mission is to steal that mojo back and the Big Data movement is going to be that next renaissance in Boston," he says. "Instead of having all these kids graduating from local schools and leaving to work for Google or Facebook, let's have them start the next Googles and Facebooks."
While Lynch has pledged to back 20 startups (and has already invested in five, including Mortar and Hadapt, both of which expand upon the Hadoop software framework that is a favorite of the Big Data crowd), he says the Boston area could probably support 100 Big Data companies.
Bob Brown tracks network research in his Alpha Doggs blog and Facebook page, as well on Twitter and Google +.
Read more about data center in Network World's Data Center section.