The investment dates back to October and was part of a Series B round that included Mayfield Fund and was led by Sequoia Capital. Versa officials say they will divulge the details of the Series B when the company officially unveils its products “in a few weeks.”
In 2012, Versa raised $14.4 million in its Series A, exclusively with Sequoia Capital, according to an SEC filing.
Vijay Doradla, a director with Verizon Ventures, says the founders’ track record was a big reason why Verizon wanted to invest in Versa. Brothers Kumar Mehta and Apurva Mehta were key figures at Juniper, where Kumar was VP of engineering and Apurva was CTO of Juniper’s mobility business unit. They led the development of the MobileNext product family — Juniper Networks’ entry into evolved packet core (EPC).
“These are not ordinary engineers,” Doradla says. “They are highly talented engineers with a phenomenal track record. Verizon Ventures looks at the business model, the strategic angle, how the investment benefits Verizon — but more importantly, we want to work with the best in the industry.”
In fact, the carrier “wanted to invest much more than we could take,” Kumar Mehta says.
“More would have been better,” acknowledges Doradla. “The amount of capital is not tied up with the amount of work we do with our portfolio companies. We like to take a hands-on partnership role and be a champion for the company.”
3 NFV Use Cases
When the Mehta brothers were at Juniper, they noted that service providers wanted to deploy services faster but were constrained by the complexity of their networks. So in 2012, the brothers decided to branch out on their own and form Versa to use NFV and software defined networking (SDN) to help service providers compete with the likes of Google and Amazon.
Versa promises that later this fall, it will clearly divulge what it’s working on. Meanwhile, the company is staying quiet.
Kumar Mehta did tell SDxCentral, “We are focused on the next-generation branch, essentially three use cases.” Those being:
“All use cases are not going down the NFV path,” says Mehta. “The core router is going to stay the way it is. But anything from the edge to access to branch is now going to be re-looked-at by service providers in terms of reducing costs. It’s a ground-up design with NFV in mind.
“We had a huge service provider who visited us, probably No. 2 in the world; they wanted a 1,000-branch network,” he adds. “Right now, it would have to take months; it would be highly complex.” He says Versa’s technology would allow that service provider to create its 1,000 branches of connects combined with security in a matter of minutes.
As for a ground-up technology and how that might work with existing network equipment, Mehta says physical network functions will “definitely not be immune from service providers [wanting] to provide an agile infrastructure.” He says the equipment vendors will have to provide features such as OpenFlow programmability in their gear.
Doradla says Versa will have to work closely with incumbents’ existing equipment vendors and that each carrier’s network will look very different from each other.
The NFV vendor field is pretty crowded, Doradla says. There are the large equipment makers such as Juniper and Alcatel-Lucent, and there are 15 to 20 startups in the space, with more coming.
“We believe Versa is probably the top of the bunch,” he says. “In five years, I think we’ll see that about 20 percent of the network has gotten virtualized. In five years, the market will be concentrated in terms of leaders.”