Jay Parikh is in charge of the back-end software and hardware that helps more than a billion people share photos and updates about their lives on Facebook.
Sounds like a nightmare. Luckily, the vice president of infrastructure engineering seems to enjoy his work. He was in town last week to do some recruiting and spend time with the company’s Boston-area engineers. Facebook’s Kendall Square office, which opened in 2013, is still relatively small, but it’s on track to grow to about 50 people. On his visit, Parikh managed to avoid the traffic problems of Harvard and MIT graduations, but did get stuck with some vintage New England rain.
Parikh is a veteran of Akamai from the early days (1999-2007). He spent a couple of years in Boston before settling in the Bay Area and then working at Marc Andreessen’s old startup, Ning. He’s been with Facebook (NASDAQ: [[ticker:FB]]) for the past six years.
He draws three parallels between Akamai and Facebook. None of them is particularly surprising, perhaps, but given how different the two businesses are, there are interesting similarities: hyper growth in the early years; a culture of moving fast; and “hiring the best people possible and empowering them”—a key criterion for “everyone walking in the door,” Parikh says.
Perhaps the most important factor now, even as Facebook has grown to 10,000-plus employees, is the focus on speed. “The faster you learn, understand the data, how things are being consumed, and the faster you can iterate, the more successful you are likely to be,” Parikh says. With the rise of mobile technology, big data analytics, and more powerful processors, he adds, “decisions need to be made much faster. The instructions per cycle are going up.”
Parikh’s time at Facebook has been marked by two “transformational phases,” as he puts it. The first was the company’s decision to design its own infrastructure—data centers, servers, and software—and open it up to the industry to get the designs more broadly adopted (the Open Compute Project). Thanks to that move, Parikh says, “We have more control over our own destiny.”
The second transformation was the shift to mobile. In the past few years, hundreds of Facebook engineers have had to learn the iOS and Android operating systems and programming, Parikh says. Facebook has made huge acquisitions in the sector: Instagram ($1 billion, which seems like a bargain now) and WhatsApp ($20 billion-plus, maybe not a bargain). Now roughly 70 percent of the company’s advertising revenues come from mobile. “Three years ago, we were at zero,” Parikh says.
Like all big companies, Facebook has to balance its need for speed with longer-term projects. The company has three long-range areas that it’s working on (and can talk about publicly): Connectivity Lab, the year-old R&D effort to provide Internet access to underdeveloped areas using drones, satellites, and other technology; artificial intelligence and “deep learning” (which is rapidly becoming the new “big data” in marketing-speak); and virtual-reality interfaces.
That last one caught my ear because I’ve been thinking about what the next mainstream computing interface might look like. “In the 5-10 year time frame, we believe there’ll be another shift in the computing platforms,” Parikh says. “What’s next after smartphones” will involve virtual reality and augmented reality, he says. His team is working on designing standards for how such platforms could operate, particularly for social interactions and experiences.
And that’s where things get interesting. Facebook’s acquisition of VR-headset startup Oculus Rift is part of the strategy here, but only part. Parikh gives some familiar scenarios that could come alive with virtual-reality tech: things like interacting with doctors, learning in virtual classrooms, and “seeing the Eiffel tower from your living room in Cambridge.”
To make that work will require innovation in both software and hardware, he says. Imagine billions of people connecting to the network with VR devices and interacting with games, content, and each other (hello, Matrix). To Parikh, the infrastructure needed to support that “will push on the same things” his team has already been tackling: namely, how do you “store more stuff, serve more stuff, compute more stuff.”
It all sounds surprisingly straightforward. But if Facebook and other companies succeed at bringing virtual reality to the masses—particularly in a mobile, portable form—get ready for more big social changes. It’s hard to believe that only 10 years ago, people weren’t glued to their phones. What will we be glued to in the next 10?
