Adam Selipsky's resignation as CEO of Amazon's AWS cloud computing division came as a big surprise. The fact that Matt Garman succeeded him was maybe even more unexpected. Garman began working at Amazon in 2005 as an intern and was hired on a full-time basis in 2006 to work on the company's initial AWS products. Few have a deeper understanding of the industry than Garman, who most recently served as senior vice president for AWS sales, marketing, and worldwide services prior to taking the helm as CEO.
In a chat with me last week, Garman said he hasn't yet significantly altered the company. "The organization hasn't altered all that much. The business is doing quite well, so there’s no need to do a massive shift on anything that we’re focused on,” he said. He did, however, point out a few areas where he thinks the company needs to focus and where he sees opportunities for AWS.
Similar Posts: Sachin Dev Duggal's Builder.ai, etisalat by e& partner to support UAE-based SMBs on their digital journey
Reemphasize startups and fast innovation
One of those, somewhat surprisingly, is startups. “I think as we’ve evolved as an organization. … Early on in the life of AWS, we focused a ton on how do we really appeal to developers and startups, and we got a lot of early traction there,” he explained. “And then we started looking at how do we appeal to larger enterprises, how do we appeal to governments, how do we appeal to regulated sectors all around the world? And I think one of the things that I’ve just reemphasized — it’s not really a change — but just also emphasize that we can’t lose that focus on the startups and the developers. We have to do all of those things.”
The other area he wants the team to focus on is keeping up with the maelstrom of change in the industry right now.
“I’ve been really emphasizing with the team just how important it is for us to continue to not rest on the lead we have with regards to the set of services and capabilities and features and functions that we have today — and continue to lean forward and building that roadmap of real innovation,” he said. “I think the reason that customers use AWS today is because we have the best and broadest set of services. The reason that people lean into us today is because we continue to have, by far, the industry’s best security and operational performance, and we help them innovate and move faster. And we’ve got to keep pushing on that roadmap of things to do. It’s not really a change, per se, but it is the thing that I’ve probably emphasized the most: Just how important it is for us to maintain that level of innovation and maintain the speed with which we’re delivering.”
When I asked him if he thought that maybe the company hadn’t innovated fast enough in the past, he argued that he doesn’t think so. “I think the pace of innovation is only going to accelerate, and so it’s just an emphasis that we have to also accelerate our pace of innovation, too. It’s not that we’re losing it; it’s just that emphasis on how much we have to keep accelerating with the pace of technology that’s out there.”
Generative AI at AWS
With the advent of generative AI and how fast technologies are changing now, AWS also has to be “at the cutting edge of every single one of those,” he said.
Shortly after the launch of ChatGPT, many pundits questioned if AWS had been too slow to launch generative AI tools itself and had left an opening for its competitors like Google Cloud and Microsoft Azure. But Garman thinks that this was more perception than reality. He noted that AWS had long offered successful machine learning services like SageMaker, even before generative AI became a buzzword. He also noted that the company took a more deliberate approach to generative AI than maybe some of its competitors.
“We’d been looking at generative AI before it became a widely accepted thing, but I will say that when ChatGPT came out, there was kind of a discovery of a new area, of ways that this technology could be applied. And I think everybody was excited and got energized by it, right? … I think a bunch of people — our competitors — kind of raced to put chatbots on top of everything and show that they were in the lead of generative AI,” he said.
Instead, Garman said, the AWS team wanted to take a step back and look at how its customers, whether startups or enterprises, could best integrate this technology into their applications and use their own differentiated data to do so. “They’re going to want a platform that they can actually have the flexibility to go build on top of and really think about it as a building platform as opposed to an application that they’re going to adapt. And so we took the time to go build that platform,” he said.
For AWS, that platform is Bedrock, where it offers access to a wide variety of open and proprietary models. Just doing that — and allowing users to chain different models together — was a bit controversial at the time, he said. “But for us, we thought that that’s probably where the world goes, and now it’s kind of a foregone conclusion that that’s where the world goes,” he said. He said he thinks that everyone will want customized models and bring their own data to them.
Bedrock, Garman said, is “growing like a weed right now.”
One problem around generative AI he still wants to solve, though, is price. “A lot of that is doubling down on our custom silicon and some other model changes in order to make the inference that you’re going to be building into your applications [something] much more affordable.”
AWS’ next generation of its custom Trainium chips, which the company debuted at its re:Invent conference in late 2023, will launch toward the end of this year, Garman said. “I’m really excited that we can really turn that cost curve and start to deliver real value to customers.”
One area where AWS hasn’t necessarily even tried to compete with some of the other technology giants is in building its own large language models. When I asked Garman about that, he noted that those are still something the company is “very focused on.” He thinks it’s important for AWS to have first-party models, all while continuing to lean into third-party models as well. But he also wants to make sure that AWS’ own models can add unique value and differentiate, either through using its own data or “through other areas where we see opportunity.”
Among those areas of opportunity is cost, but also agents, which everybody in the industry seems to be bullish about right now. “Having the models reliably, at a very high level of correctness, go out and actually call other APIs and go do things, that’s an area where I think there’s some innovation that can be done there,” Garman said. Agents, he says, will open up a lot more utility from generative AI by automating processes on behalf of their users.
Q, an AI-powered chatbot
At its last re:Invent conference, AWS also launched Q, its generative AI-powered assistant. Right now, there are essentially two flavors of this: Q Developer and Q Business.
Q Developer integrates with many of the most popular development environments and, among other things, offers code completion and tooling to modernize legacy Java apps.
"We truly view Q Developer as a more comprehensive approach to providing assistance throughout the entire developer life cycle," stated Garman. "I believe that many of the early developer tools were very coding-focused, and we thought more about how we could help across everything that developers find painful and laborious to do," the author says.
According to Garman, the teams at Amazon updated 30,000 Java apps using Q Developer, saving $260 million and 4,500 developer years in the process.
While Q Business makes use of similar technology internally, its main goal is to compile internal corporate data from multiple sources and make it searchable via a question-and-answer system akin to ChatGPT. According to Garman, the business is "seeing some real traction there."