He spoke about the passion that AWS has for startups, and the impact startups have on the AWS ecosystem: “Startups is one of those things that I'm most passionate about, and I would say AWS is incredibly passionate about as well.

“Startups are really the foundation at AWS. They are a significant percentage of the companies that are running on top of AWS today, and subsequently, the global revenue that is actually generated from companies that at one point were startups on AWS. So we take that very seriously, and we see it as a responsibility to support the growth of companies throughout every stage of their maturation, particularly when we start talking about pre-seed companies, or even before taking on investment rounds. So that's a heavy focus for us.”

Bennett highlighted some recent shifts within the startup space. Traditional startups from five years ago, compared to what we are seeing today, is ‘AI-first’ startups. Startups that have been built around AI from their inception. Scaling is one of the most important things for a startup, and newer AI-first startups have the edge when it comes to scaling. “We hear very often from companies that have to be very thoughtful about how they can really grow effectively and how they can scale quickly. We find that to be even more so the case with AI-first startups, because these companies can scale in a matter of months, in a matter of quarters, really exponentially,” explained Bennett. “These companies often need to have access to enterprise capabilities, tools, and services from day one, and a part of that is because they can't wait six years to get the things that they need from a service perspective, from a marketing perspective. There’s a bunch of capabilities that they expect day one, and so we have to engage with those companies a little bit differently.”

AWS supports over 300,000 startups built on top of AWS with services and tools that help them focus on their business and differentiation, across every industry vertical. They're solving unique and complex problems that are can be very bespoke and specific to a particular industry, but they have very broad aspirations.

Looking at how AWS interacts and supports these startups, Bennett explained: “Our focus on the startup segment is really not just around the volume of startups, but it's about how we support them broadly, particularly as they're growing. That's part of the reason why AWS has supported startups with over $7 billion in credits and investments throughout the years. And it's really important to note that it wasn't just that we provided credits and support. We also brought a lot of our resources. We have a number of teams inside of AWS. If you look at our startup business inside of AWS, you will actually find a number of founders, CTOs, a number of folks that have lived the experience of being at a startup, and we bring that to our customers, we bring that support and insight.

“And for many of our startups, it's not just that they're thinking about how they're building great tools, products, solutions, or experiences for their customers, they're also thinking about how they build a viable and sustainable business long term. And that often means thinking about what go to market will look like, thinking about how they grow different parts of their organisation to be effective over a period of time. So those are just some of the areas that are top of mind for us as we think about this critical space for startups.”

AWS Generative AI Accelerator and AI-first startups

Whilst it seems like AI has been overtaking and people are looking for the next new thing, in reality, it has only been in the last few years that this has been the case. Bennett was asked what he saw as the next big trend moving away from AI, but expressed that predicting trends has become difficult, and not only that, but how only a couple of years ago, few people were even discussing generative AI in detail.

The rapid emergence of generative AI, agentic systems, and deep-reasoning models show how quickly the landscape is shifting. He believes value creation still sits across a wide range of areas, with opportunities in agentic orchestration for the enterprise and in improving how customers engage with large language models.

Looking ahead, he said investors expect faster execution from startups, which influences funding behaviour, including smaller seed rounds to accelerate development, and how falling costs and greater efficiency means more companies might choose to bootstrap for longer, which is currently an important trend for the startup ecosystem.

When it comes to AI-first companies, there are a wide array of startups that fit into this bracket. A hot topic of AWS re:Invent was agentic AI, which is one of these categories, however there is also generative AI and physical AI startups making waves in the space.

“There are a very broad range of companies in our Generative AI Accelerator (GAIA). We’ve intentionally leaned in with a number of companies. We provide a significant amount of credits and services and resources to actually help those companies accelerate their development cycles and build effectively with the intention of being able to deliver a product that their customers will love even faster.”

However, GAIA is not the only programme that AWS offers for these companies. It has also niched down and offers fellowship programmes for robotics and physical AI startups, where it provides similar support and investment models with a number of customers.

What is fundable right now?

At an ‘ask me anything’ session, Bennett recalled an early experience of building an enterprise back in 2016-17, and the lessons learned. He explained that predicting which customers would adopt the Cloud eventually proved unreliable because circumstances consistently shift. Leadership changed, boards changed, and wider economic and geopolitical factors altered priorities of customers.

Having been asked about the current landscape of AI startups, he believes this experience can be applied to startups. Entry barriers to found a startup have fallen, and it has become easier to build something, which makes it harder to judge early on which ventures would be the most suited to invest in. However, AWS continues to support all startups, and works closely with venture capital firms, universities, and accelerator programmes, to identify opportunities when potential is spotted.

Listening to customers

He went on to discuss what AWS provides, and how these products and services came to be. AWS continues to be a company that evolves and its offerings have become broader.

Bennett said: “We want to help companies be successful quickly … For AWS, roughly 90% of what we build, the services, the features, it's actually built based on the insights and the inputs that we get from our customers. The other 10% is from us, inferring what we think our customers may be asking for, but maybe what doesn't exist today. A lot of value comes from driving that level of support and engagement with those end customers, because it tells us a lot about where we should go and build for the future.”

He also discussed how some of the launches announced at the event, such as autonomous agents with KIRO, to Nova Forge, will impact startups building the future. “You can imagine the value for many startups. There’s two aspects that they've been looking at previously. The first one is, I leverage a pre-trained model, and then I bring my data to it. The other is, I'm going to build my own foundational model from scratch.

“Those are two very different pathways to go down, and the implications from a capital perspective around both of those are pretty unique as well. So we're pretty excited around Nova Forge, because we had enough customers that were asking us and saying, ‘is there a way that I can bring a lot of that insight and a lot of that data at the right point in time to actually build my own model, and then if I can, how can I leverage that effectively? How can I leverage that to drive a lot of the agentic workflows that I'm actually thinking about more private?’”

The work that AWS is putting in to support its customers, from enterprises to startups, shows that it is committed to being a supporting force with them from day one.

For more startup news, check out the other articles on the website, and subscribe to the magazine for free. Listen to The Cereal Entrepreneur podcast for more interviews with entrepreneurs and big-hitters in the startup ecosystem.