Ten years ago, Justin Kan and Emmett Shear had just sold their app company, Kiko, and were itching for another venture. They had a concept — livestream video — but no idea how to build it. So they sent an email to the MIT engineering listserv, requesting a “hardware hacker” for an unspecified project.
Kyle Vogt, a young student fascinated with robotics, replied. They met over coffee where Kan and Shear pitched their idea before flying out to San Francisco. When the pair landed, they had an email from Vogt with a dense PDF attached — he had drawn out the full hardware specifications for their theoretical live-streaming camera.
“This guy is on the ball,” Kan recalled thinking. “We need to bring him on.”
They did. Vogt would lead product for their startup, Justin.tv, then go on to start its descendant, Twitch, with Shear. Then he started another company, Cruise Automation. In under three years, Vogt achieved what few tech founders even dare to dream: Twitch, a video-streaming service for gamers, sold to Amazon for just under $1 billion; on Friday, Cruise, an autonomous driving startup, sold to General Motors for even more.
Vogt is the Silicon Valley’s latest entrepreneur with the Midas touch, a favored founder with its monied class as tech begins to collide with colossal established industries, like cars.
“At Cruise, he was the only one for three months to build it from nothing to a prototype, both the hardware and software. That is a rare level of talent,” said Sam Altman, head of the Y Combinator startup incubator. “I can see Kyle being the next CEO of GM.”
Timing helped Vogt, too. The engineer, who is 30, worked on two technologies — livestreaming video and self-driving cars — just as they landed on the radar of big, deep-pocketed companies. But people who know Vogt attribute his success to a rare mixture of technical know-how and business acumen, along with an obsession with difficult feats.
“Not only is he not afraid of hard technical problems, but he seeks them out,” said Michael Seibel, a former colleague and partner at the Y Combinator incubator.
He is certainly not the only techie working on the hard problem of building cars that drive themselves. What sets Vogt apart is a fastidious attention to the logistics of how people — and companies — will interact with these cars, said Nabeel Hyatt, an investor with Spark Capital, which led Cruise’s first financing round.
“That takes a polyglot,” he said. “If an engineer is slowing down, he’s not afraid to write a line of code. But at the high level he is thinking about marketing, brand, consumer experience.”
Former colleagues and investors spoke of Vogt’s longstanding obsession with robotics. He was part of the early self-driving car experiments at MIT, interned at Roomba-maker iRobot and competed in two seasons of BattleBots. (At Justin.tv, Vogt hatched a plan for a beer-delivery robot, but it never materialized.) “He’s always been a robot guru,” said Kan.
In the fall of 2013, Vogt left Twitch to start up his own company. (Amazon bought Twitch a year later.) He linked up with Kan’s brother, Daniel, who had met Vogt at Justin.tv. And they brought their idea — a system for retrofitting cars with self-driving tech — to Y Combinator, an incubator known for software companies like Airbnb and Dropbox, not hardware.
Y Combinator hadn’t done cars; Cruise was their first incubation in the industry. But the partners knew Vogt. And they saw his pitch as akin to what Justin.tv and Twitch did, building a video-streaming system from scratch. “It seemed like the exact same thing: Let us take on a problem that no one has solved,” said Seibel.
Within a year, the Cruise team had a rigged-up prototype. They took the incubator’s chiefs down the 101 freeway in Northern California. “It was pretty ghetto,” said Justin Kan, now a Y Combinator partner.
But it drove. “I went in it and offered to invest on the spot,” said Altman.
Vogt also displayed an aptitude for poaching talent; in September, Cruise hired Andrew Gray, a lead engineer on Tesla’s autopilot software, as its VP, a move that made a splash in automotive robotics circles.
The startup was buzzy in that insular world before then, though its business model was never clear to outsiders. Nor was it entirely clear to investors. (One sign of Vogt’s reverence in Silicon Valley: Cruise’s investor list reads like a Who’s Who of tech; however, several of them were in the dark about Cruise’s plans, including the GM acquisition.) The company started building after-market kits that retrofitted cars with autopilot features, partnering with Audi.
At some point last year, Vogt shifted plans to focus on building a full self-driving platform. That strategy pits the startup against Google, which is, according to multiple industry sources, pitching a similar system to carmakers. As a part of GM, Vogt will now have many more vehicles at his disposal — and will face far more intense strain testing his hardware hacking against the largest tech companies in the world.
It’s tough to imagine Vogt, who has spent his career buzzing from one startup to the next, will linger at the 98-year-old Detroit carmaker. Yet those close to him insisted that he is dedicated to staying at the company. GM probably structured the deal that way, if the company is smart.
“Google may still be a formidable competitor,” said Altman. “But there is a reason startups only doing one thing can often win, again and again.”
Vogt declined to comment for the story. On Friday morning, he simply sent out a perfunctory tweet.
Additional reporting by Kara Swisher.
This article originally appeared on Recode.net.