Sramana Mitra: What year did you quit Borland?
Patrick Kerpan: It was sometime in early 2006. I was going to do an excellent year off. After about two weeks of me reorganizing household processes, my wife said, “You’ve got to get an office outside of the house because you’re driving us all crazy.” The co-founder of the company and the Chairman called me and wanted me to do due diligence on some company in London. One of our other founders introduced me to Alexis Richardson. Our former head of security had gone in to help a friend cleaned up and get out of high-frequency trading.
As a group of people who’re experienced, it was like the old movies where they’re like, “Let’s put on a show. My mom will make costumes.” We wanted to build a company, but we had no idea what it would do. We spent the better part of 2006 focusing on what we would be doing. We came back together around mid-year and said, “Open source, open standards, loosely coupled architectures, and virtualization are going to change the way data center computing is done.” That was our Aha. Then we said, “With that as the context, what do we do?” So we started putting together different business plans and floating them around with friends and families, Swiss Bank, and angel networks.
We finally settled on a plan for building a platform for virtual machine construction. Dell has a factory where you can customize and say, “I want this much memory and this much disk. I want it to be blue.” At that time, Amazon was just coming out. This was really targeting the early virtualization world of VMWare, Zen, Virtual Iron and let you build a virtual computer that would then run in any of these different environments. It was a fascinating service.
Our go-to market model was, “We know a lot of people in the financial industry. This is the kind of thing that they need from an audit and compliance point of view.” Pretty quickly, the financial crisis hit. Half the people we had known didn’t have jobs anymore. Global capital market companies weren’t buying. We moved forward a little bit too long.
We had built a second service to secure our service. This was Cohesive Networks. What I’m talking about is Cohesive Flexible Technologies. When we first brought this service to market, Amazon wasn’t there. The first time that we went to spread, it was Tier 1 Dallas and Tier 1 Los Angeles. We went, “How does all our data move off the racks and across Tier 1 and over the Internet safely? We went to the web browser and we thought that somebody would have already done this. We were calling it an overlay network. There’s got to be a way that you just use software and you just build a network out of software. It’s encrypted end-to-end. The Tier computers are sitting on the actual networks. It makes it look like it’s one logical address so you don’t have to reconfigure anything.
We just naively thought that had to exist. We couldn’t find anything, so we wrote this cheap and cheerful thing to secure our service. Fast forward another six months or a year, every customer we ran into actually needed this network virtualization piece for their application running on top of the cloud more than they wanted our image automation engine. The go-to market was very different. For a small lightly capitalized company, I think you have to find a way to have a product that people will buy. You can’t have to sell. Part of our product line required selling and part of our product line was just bought. At that time, I was the CTO and Craig was the CEO. We doubled down on enterprise.
Now, it’s 2011. We spent a lot of time with IBM and IBM SmartCloud on using the whole thing as a solution for migrating from enterprise data center to cloud. We got some decent revenues. We were still probably burning a dollar for every 50 cents we made. You have this spiky pipeline. You’ve got one-off revenue coming in and it’s not necessarily reproducible. So we made me CEO in 2012. It is now Cohesive Networks where you buy subscription to virtual networks.
The virtual networks are created out of virtual machines that the customer controls. It’s really like a virtual router, switch, and firewall. It now fits in that world of that software-defined networking and then network function virtualization. Nobody called it that when we built it. In 2013 we went to the Board and said, “We’re killing everything else. We’re not going to take any revenue. We’re all networking.”
This segment is part 3 in the series : Successful Pivot to $5M in Revenue from Chicago: Cohesive Networks CEO Patrick Kerpan
1 2 3 4 5 6 7