I worked with Jeff Bezos at Amazon in the early 2000s. He was amazingly intense and made me rethink the way I view the world.

• David Selinger cofounded Deep Sentinel, an AI security firm, after working at Amazon in the early 2000s.
• At Amazon, Selinger worked directly with Jeff Bezos, which shaped his AI and business acumen.
• Deep Sentinel, backed by Bezos, uses AI to prevent crime and innovate home security.

This as-told-to essay is based on a conversation with David Selinger, the 48-year-old CEO of Deep Sentinel, based in Pleasanton, California. It's been edited for length and clarity.

I went to Stanford but didn't finish my degree, so I went back in 2000 to finish. In 2002, I got hired at Amazon through on-campus interviews while still in college.

I worked at Dutch Bros, now a publicly traded coffee company, then. I was running their tech side.

I didn't initially want to work at Amazon, but I considered it because I was bored. I had a great job at Dutch Bros, worked very few hours, and made a ton of money. I took the Amazon interview very nonchalantly.

I found the data side of Amazon appealing

At Stanford, I studied machine learning, robotics, and AI, but Amazon's dataset was unique. Few companies at the time used the word "terabyte" to describe data.

Amazon sucked me in because of the opportunity to work closely with Jeff and to work on something that I wasn't very experienced with, which was large data sets. I got the job.

My job title was the manager of customer behavior research

I joined what became the customer behavior research group at Amazon in January 2003. At that point, we had a few thousand employees, and we were in an office called Pac Med. On average, I met with Jeff once or twice a week.

My office was down the hall from Jeff, and I could hear him laughing constantly. He was great and amazingly intense; I greatly enjoyed working with him. Jeff Bezos made me rethink the way I view the world.

I worked 80 hours a week at Amazon in a gladiatorial culture. You got stuff done. Amazon removed the coordination and restrictions that other traditional companies had. Everyone criticized Jeff for that, but that's also what motivated people like me to do things.

Jeff saw e-commerce as a trillion-dollar opportunity before it became mainstream

He had the vision. Jeff believed, without exception or a willingness to compromise, that nobody else's opinion mattered about how hard or how impossible something was, that something could truly be that big.

When I joined Amazon, the media was saying Amazon would go out of business this quarter or next — that was the assumption. Jeff steamrolled ahead, and Amazon took on debt to build warehouses because he had confidence and clarity, which enabled him to overcome what everyone else said was impossible.

Jeff taught me that common sense requires common assumptions

One of the things Jeff invented was the two-pizza teams. It was like an org chart, with little teams led by essentially Rambo squads — each operates independently. They don't collaborate; they're sent out to battle.

Many people said this would fail. That was all based on common-sense knowledge — if you do that, people will quit. Jeff said, "Okay, let's deal with the assumption. Assumption No. 1 is that I want to keep people." Businesses assume the leaders want to keep people.

In its early days, Amazon often rotated leadership out of mid-management every 12 to 24 months. Jeff believed he could attract the best talent and get the most out of them in those 24 months.

Amazon reinvented the idea of an organizational structure, from being for stability to being for volatility, and discovered the greatest talent in that way.

Jeff taught me that data can transform anything

That wasn't obvious in 2000, 2002, or even 1995. When I built my team, Jeff had to convince the board of directors that building a team that would use Amazon data was unique.

My biggest accomplishment at Amazon was convincing Jeff to allow the advertising business unit to exist and grow. In our first meeting about the potential of ads on Amazon, Jeff said that Amazon is a retailer — we don't put ads on our website. After I showed him the data that proved we should run ads, he approved them.

I was recognized as a high performer, but I still decided to leave

I walked away from so much stock. At the time, I didn't realize it would be that big, but I wanted to start my own thing.

I left Amazon in 2004 and cofounded Redfin, a real estate company, and served as its CTO. I was full-time for a year, got it funded, got them to hire a CEO, and then I moved on.

I then started Rich Relevance, a machine learning enterprise software company for big retailers, which would eventually merge with another company. I was in the hopping stage of my life.

I next started an AI security company called Deep Sentinel

Throughout my career, my love for AI has grown. I wanted to start a company that would have an unequivocal positive impact on the world.

Deep Sentinel provides a security solution that includes cameras and a back-end service that connects the cameras to an AI and connects that AI to live guards. We stop thousands of active crimes every month at an affordable price point.

We manufacture our cameras and provide the guard service, and I oversee and manage it. We built all the software and all the back end ourselves.

I self-funded the start of Deep Sentinel and invested about $1M in it. Then, we funded it in partnership with Jeff Bezos, one of our earliest investors. I reached out to Jeff, and he agreed to be a part of it.

My daughters helped me get Deep Sentinel started

They did some of the early branding work and logos, and talked about how it would work and operate, and what technology we'd build.

I'm a dad and a CEO. I still work a lot. I have offices in Makati, Philippines, Taiwan, and Pleasanton, California. All my employees in the US are remote. I go to the office once or twice a month.

My experience at Amazon became the foundation for my career.