Theophilus "Theo" A. Benson is joining Brown CS as Assistant Professor in September, and if you see him out on a run, training for his next Tough Mudder, consider joining him for at least a block or two. He’s always eager to discuss recent IoT security challenges, cloud outages, and performance issues.
Theo’s introduction to CS dates back to the era when cloud computing was only possible in science fiction: “I remember there not being a lot of color and complexity on the computer screen. Back then, it was so much simpler to find out how things worked.” His interest in the field continued through high school, but an early job at a startup, doing testing and development, systems engineering, and release engineering, provided additional momentum and a definite direction for Theo’s career. “One day,” he remembers, “I got paged because the sysadmin was out sick and the server had gone down. I understood the release process, so I thought I’d just reboot and things would be fine. The problem was that some of the process had been documented, but part hadn’t, and a lot had to be inferred. Not quite as easy as I’d thought!”
“And that got me thinking about the broad space of networking,” he says, “and how to take humans out of the loop when they’re going to cause errors or slow things down. The problem is, we don’t want to be taken out of it!”
In graduate school at University of Wisconsin, Theo focused his attention on the intersection of networks and security, using the tool of configuration management, a process that analyzes the requirements, design, and performance of a network to improve the consistency of its performance. “Configuration management is a huge challenge due to competing economic interests,” he says. “There can be a lot of animosity when sysadmins think that we’re taking away their jobs, but we’re really just trying to minimize the time they spend on certain tasks.” He gives the analogy of going to the doctor: patients describe their symptoms but don’t diagnose their own illnesses.
Despite industry experience that also includes AT&T and Microsoft, academia has always felt like home: “There’s a lot more freedom to attack problems. In the corporate world, when a new buzzword arrives, everyone switches over because it’s the next big thing. I like the wide scope of interests at Brown CS. Instead of being in a silo of networking experts, I’m surrounded by different perspectives.”
When we ask Theo about the outreach efforts listed on his web site (among other things, his research group has allowed high school students to collaborate with them during the summer), his answer is refreshingly free of platitudes. “That experience was very valuable to the students,” he says, “but I don’t feel like I’m being helpful enough. As long as we’re interacting with students at their level, it helps, but getting outreach right isn’t easy, and I’m still trying to figure things out. A lot has been given to me, so I feel like I need to do more.”
He’ll have a chance in Providence, where he’s looking forward to being physically closer to both family and colleagues in the area. Part of the Brown CS appeal, he explains, is that he’ll be working alongside experts in software-defined infrastructure, programming languages, and the Internet of Things, not just networking. “We’ll amplify each other,” Theo says. “And I’ve heard great things about the undergrads and how they take a huge amount of initiative in jumping in with research. I want to see what we can do together.”
Some of the research possibilities include the area of software-defined networking, which he sees as going through a major paradigm shift. “There’s going to be a whole new level of automation,” he says. “Instead of transplanting network protocols designed for Google to developing countries, where there are huge infrastructure discrepancies, we’ll have algorithms that can learn about local conditions and adjust, networks that are able to learn by themselves.”
“It requires us,” explains Theo, “to question assumptions about what works and what doesn’t.” It’s part of what appeals to him about the work of Ion Stoica, whom he cites as one of his heroes. “I always looked up to his work with Spark, which really impacted the way big data interact with networks.” In the end, he says, meaningful computer science speaks for itself far better than becoming a household name. “When you have that much impact, that’s what people remember. They may not know your name, but if it works, it works!”