After 23 years working for Jeff Bezos, the CEO of a $1.3 billion skills platform shares lessons he learned from Andy Jassy and the Amazon founder

In 1997, the day before Greg Hart joined Amazon, he was summoned to a meeting—on a Sunday—with its founder, Jeff Bezos.

At the time, Bezos had interviewed virtually every one of Amazon’s circa 200 employees; Hart was one of the few the tech entrepreneur hadn’t personally appointed. Over the next 23 years at the online giant, Hart reported directly to Bezos as technical advisor to the CEO, and to Amazon’s current CEO, Andy Jassy.

The lessons Hart learned at one of the world’s most famous businesses have stayed with him to this day, where he leads $1.35 billion online learning giant, Coursera. Hart tasked himself with shepherding the company through a transformation—conveniently, in time for demand to explode, as job seekers and employees alike rushed to add an all-important AI qualification to their CVs.

Many of the changes Hart brought to Coursera—and its more than 1,000 employees—will be familiar to Amazon alumni. Hart said that Bezos’ practice of interviewing every employee in the early days set the tone as Amazon grew, explaining, “He wanted to make sure that the passion, customer focus, the high standards, and the move fast traits that the early set of employees had stayed true as the company grew in scale.”

So it made “perfect sense” when Bezos penned his now-famous letter to shareholders outlining the leadership principles and priorities of the business, because they “reflected” the day-to-day conversations in the office.

Hart wanted to embed a similar mindset at Coursera, he said: “I wanted to really transform the company and make it move at a faster rate and do a better job of serving our learners. I felt that one of the most critical things in doing that was ensuring there was really good cultural alignment, and so we introduced a set of leadership mindsets. We looked at some of the most successful companies in the world, we looked at either their values or their principles … and we created our own that we felt were very specific to both our business and our history as a company.”

That speed became critical as the AI boom transformed the skillset businesses wanted, with employees and job seekers racing to keep up. The platform is now home to more than 12,000 courses, 1,100 of which are based on generative AI—a 44% increase year-over-year. GenAI is significantly the most popular topic on the platform, both from individual learners and from employees with a subscription paid for by their employer.

The CEO was also keen to do away with unfocused company all-hands, and instead dusted off the Amazon playbook of focusing each of the meetings on a single leadership principle: “One of the things that I’ve just recognized over my time leading different businesses in different industries is no matter how clear something is in your mind, or your leadership team’s mind, you cannot repeat it frequently enough to the rest of the organization. They might not be paying attention, they may not understand it, they might have been in a customer meeting at that time, whatever, they might have missed it.

“Every month, one of my direct reports sends out an email with a video that talks about just one of our leadership mindsets. Every all-hands, we do the same thing. We’ll pick one, and have examples that speak to it, because it helps make it real for people and helps people have better context around it.”

A key focus for every CEO at present is how they can leverage AI at work, either within their business or in their own personal use. KPMG’s 2025 U.S. CEO Outlook found 74% of leaders said investing in AI was a top priority despite economic uncertainty, with 79% saying they were confident they were ahead of the curve on adoption and utilization.

Previously, CEOs have told Fortune they’re using AI for everything from recruitment to management, to meeting prep and document summaries.

Hart, an English major, is well-versed in the efficiencies AI can offer but said one thing he never uses the technology for is writing. “For me, writing is the way I think, and so trying to outsource that would be effectively giving up thinking,” Hart said. “So that would not be appealing or effective for me personally.”

Staff across Coursera are encouraged to experiment with AI as they see fit, currently without any goalposts in place for what they should be trying to achieve. The most useful outcome of this approach, Hart adds, is that colleagues are sharing their use-cases and best practices in an internal forum called ‘AI Sparks.’

“AI Sparks is a monthly meeting where people from across the company, at any level, are coming to share how they’re using AI in their job. Those are by far the most well-attended and popular meetings that we do at the company,” Hart said.

A final lesson from Amazon prepared Hart for the era of AI: If you get too caught up on outcomes in the early stages of a new technology, you miss the bigger picture.

“My perspective is we just want to get a workforce that is using it as much as possible, in as many ways as possible. Over time, we’ll start to be much more focused on quantifying the impact on all of that,” Hart said. “If we focus myopically on that right now, I think we would miss the opportunity to have a far greater impact down the line.”

This story was originally featured on Fortune.com