Building Your Own Cloud Is “Table Stakes,” Says Former AWS Engineer

What the heck is Jesse Robbins doing starting a next-generation walkie-talkie company?

Robbins, the founder and CEO of Orion Labs, a company that builds a wearable communication accessory for instant voice conversations, is better known as the founder of Chef, an IT automation startup beloved by DevOps engineers everywhere. Before that, he was part of Amazon’s infrastructure team, one of the earliest engineers to work on AWS. 

See also: How To Choose A DBaaS For Database Management

He’s an infrastructure guy. A software guy. But a walkie-talkie guy? 

Turns out, Orion sits at the heart of a personal Venn diagram for Robbins, an opportunity to mix his infrastructure expertise with his volunteer firefighter work (yes, really). 

But ultimately, it’s a lesson for every company: We’re all in the infrastructure business now.

A Long Day’s Journey Into Wearables

When I sat down to talk with Robbins, I was surprised to learn he had left Chef to start Orion. That was part of the plan, he told me. 

“In 2012 I had made the decision to leave Chef and was trying to figure out when to leave, and what to do next,” said Robbins, though he remains an active advisor to Chef. “We wanted to manage the process carefully.” 

In addition to being a firefighter, Robbins is also a sailor, and on a long sail, he started thinking about he wanted to do next. After mulling over his options, he determined to move beyond spinning up another infrastructure company. That would be too much of the same. More importantly, he wanted to impact people directly. 

As he considered his various priorities—an infrastructure background, a desire to touch people directly, and his experience as a firefighter—he settled on real-time communications and, particularly, “stuff that I’ve used as a firefighter” like dispatch systems.

“Let’s try to build a messaging system that gets people heads-up, really focusing on the harder problem of voice and data in real-time,” he told me. Otherwise put, he was determined to build a distributed Internet of Things (IoT) platform optimized for real-time experiences. 

“It’s a very hard technology problem that has to be easy for consumers,” he details, which is what captured his attention. It can’t be solved with conventional approaches, and there’s no app for that, and can’t be. It requires thinking different.

Like, for example, the need to build out all his own infrastructure.

The Infrastructure Cometh

Apple has done it. Google has done it. Samsung has done it.

Indeed, everyone that is serious about owning the customer experience seems to be building its own infrastructure, and not just big companies, according to Robbins: 

Whether we’re early on a trend or not, we know how to build the kind of infrastructure required. There are table stakes problems that you can’t solve unless you have my team’s kind of background. We’re not a lone wolf. We’re ahead of the cycle in a spot that right now only large companies can set out to solve.

That infrastructure can be the competitive difference, Robbins insists. “A lot of people are bolding products that connect to a smartphone,” he said. “But what we do is have thousands of people connecting simultaneously” to a distributed communications network. 

In other words, “We have all the problems that Whatsapp has, but we also have to build it into a device,” he added. 

It’s a difficult problem and, notwithstanding his open source DNA, Robbins isn’t keen on sharing the know-how with others: “We don’t see a lot of advantage in explaining the internals about how we do what we do, so we don’t blog it or otherwise speak publicly about it.” 

In other words, go figure out your own distributed IoT platform. 

Beyond Infrastructure

“We’re probably an early signal,” Robbins admitted. And he’s probably right, both on the “early” part and on the “signal.”

Early, because infrastructure remains hard, and goes beyond the comfort zone of most companies, large or small. 

In the case of Orion, “It took us 12 people and 12 months to build and ship Onyx, and it required a lot of additional support resources,” Robbins said. “We were out of our comfort zone much of the time, and now we have operational expertise.” 

And that’s with a team that has a deep understanding of infrastructure engineering.

It would have been easier if he simply scaled up on someone else’s cloud. But that option wasn’t available to him, he insists. There were no shortcuts for Orion, but “the new successes will have a new set of skills,” he said. 

The industry has spent the last few years getting comfortable with building on others’ clouds, and it looks to be heading there at full throttle now. But if Robbins is correct, at some point, we’ll have to revert to operating our own infrastructure again to be competitive—and get “back to the future.” 

Lead photo by Robert Scoble 

Facebook Comments