Jeff Arnold and Waseem Daher: Ksplice
The story of Ksplice starts with four undergraduates at MIT. The students (Jeff Arnold, Ksplice’s CEO; Waseem Daher, COO; Tim Abbott, CTO; and Anders Kaseorg, Chief Architect) were involved in the student computer group and they ran servers for the MIT community. A software update was announced in the middle of the week, and Jeff decided to wait until Sunday at 3am install it so as to minimize disruption of the the system. Unfortunately, between Wednesday when the update came out and the Sunday installation the system was broken into and hackers took advantage of the problem that the update would fix, forcing Arnold and others to reinstall the system. “That got me thinking: why do we have to reboot to install updates” Jeff queried? He went on to tackle this problem in his award-winning MIT master’s thesis, which laid the groundwork for the Ksplice technology – the rebootless update. Installing updates may be an inconvenience for individuals, but it has real financial cost for servers and other IT infrastructure. “The need was very real; if we have this problem, others must too,” Waseem Daher, Ksplice’s COO, realized. Determined to solve the problem, Waseem continued, “There has to be a better way.”
Entrepreneur: Jeff Arnold (CEO)
Partners: Waseem Daher (COO), Tim Abbot (CTO), Anders Kaseorg (Chief Architect)
Company: Ksplice
Date founded: June 2008
Achievements
- 2009 MIT $100K business plan winner
- 1st place in the security category at the Trophées du Libre international software competition, 2009
- winner of the 2009 Cyber Security Challenge in the London-based Global Security Challenge
- Phase I and II National Science Foundation Small Business Innovative Research (SBIR) award (2009-2010)
Number of employees: 15
Year born: 1985
The team members all had the entrepreneurial bug but as Waseem described it, “We weren’t sure of exactly what to start; we just knew that we wanted to do a startup. And that there should be a market for this.” The Ksplice team felt that what they were doing was novel and difficult. They knew that they had developed something powerful. Waseem described his thoughts, “Even if the initial plan was not quite right we knew that we were a stone’s throw away from applicability to someone.
Background. Before founding the company, the team spent time in the field talking to others in charge of system administration. They asked questions, they listened, and they realized they were on to something that could be the basis for a startup. From a technical point of view they knew that it was a very difficult problem to solve. “The pain of bringing a system down to update it has been around forever,” Waseem stated. He continued, “The way people write software is from the understanding that the system will be rebooted and started from scratch; that’s the DNA of coding,” he explained. “Can you imagine driving your car on the road while the mechanic is working on it,” he asks?
The problem they were solving was clearly not new – others had tried to solve it. The research direction that Jeff pioneered was a different spin on the problem: “I just looked at the problem in a slightly different way.” In thinking about a company, the team leveraged the MIT community: one referral led to others. It was June of 2008; Jeff and Wasseem lined up meetings and conversations. And they got great feedback. The company was born. And they won the MIT $100K business plan competition that year.
Today. Ksplice is doing great today, with over 100,000 systems running its rebootless software and over 600 paying customers. The foursome are as tight as ever “When you have worked a lot together as students, you build a culture of trust and know that you can spend long periods of time together,” Waseem declared. Jeff, Waseem and Tim, all 25 now, were rooomates at MIT. Jeff and Waseem took an entrepreneurship class together and, together, their interest in entrepreneurship obviously has blossomed. “It’s fun to make things that people want and people use.” Waseem told me. He added, “But it raises all these other questions about business models.” They had to learn a lot of that the hard way.
While the company has received no equity funding, it has received funding and support from the National Science Foundation, with both a Phase I and II SBIR.
“Where does the Ksplice name come from,” I asked? Waseem told me the story, “The core of the offering is the kernel. And we are inserting code in the running kernel. In effect, we are splicing it in.” Go Ksplice!