Where You Build and Run Your Applications Matters – Futurum Tech Webcast

Disclaimer: The Futurum Tech Webcast is for information and entertainment purposes only. Over the course of this webcast, we may talk about companies that are publicly traded, and we may even reference that fact and their equity share price, but please do not take anything that we say as a recommendation about what you should do with your investment dollars. We are not investment advisors, and we ask that you do not treat us as such.

Transcript:

Daniel Newman: Hey everyone. Welcome back to another Futurum Tech podcast. I’m your host, Daniel Newman, CEO of The Futurum Group. Excited for this interview series today. I’m talking to AWS Tom Godden, and we’re going to be talking about where you build and run your applications and why that matters. Another great conversation with AWS, the world’s largest infrastructure as a service provider and one of the most exciting innovation engines of the cloud economy. Tom, welcome to the show. Excited to have you join me today.

Tom Godden: Yeah, I’m really excited to join you as well and to talk about this topic, something I’m obviously very passionate about. Thank you for the kind words about AWS, but we believe that we’ve earned that, that we’ve worked hard to be able to deliver and to meet the customer needs on that.

Daniel Newman: Hey, Tom, listen, I’m an analyst. I wouldn’t say it if the facts didn’t back it up, or at least I shouldn’t. And the largest is emphatically true, and if you look at the services that you deliver on a quarter to quarter basis, it’s hard to find another cloud provider doing anything of that magnitude at this moment. Now, again, I’m also a big proponent of innovation in ecosystems, so I cheer for your competitors. I cheer for you. I cheer for the idea that innovation drives innovation. And so what I do believe though, is that enterprises also need to be able to leverage, implement, and utilize all this technology and implement it quickly to get value out of it. So I actually think a great place to start though, and I’m really excited to talk to you today is just getting a little bit more of your background. I was lucky enough to have the chance to spend some time with you ahead of this podcast, and I was really blown away about you, the team, what you’re doing. Can you just give us a little bit of background what landed you at AWS and tell us about the team that you’re part of, because I think what you’re doing is really interesting.

Tom Godden: Yeah, so thank you very much, Daniel. So I am an Enterprise Strategist with Amazon Web Services. Our enterprise strategy team is a team that’s made up of 12 former Chief Information Officers, Chief Technology Officers, Chief Financial Officers from companies like Coca-Cola, McDonald’s, Capital One, NASA JPL. I was the Chief Information Officer at Foundation Medicine, the world’s largest genomics company, and I led them to go all into AWS to move from a waterfall-based methodology to an agile-based methodology, and to do that for FDA Class III regulated devices. And I found an opportunity to join this absolutely wonderful, brilliant team where we spend time with almost 2,000 customers per year, sharing not only our experience, but learning from them on what they’re doing for their digital transformations, their cloud transformations, how they’re pursuing generative AI. And we share those best practices through a variety of books and blogs, through large Keynote presentations. But really just trying to share that information and that knowledge to cut down on the learning cycle to help our customers be successful.

Daniel Newman: Yeah, it’s a huge job, and I really admire the group you’ve put together. One of the, I always call it the ground sources of truth in my business as an industry analyst and running an industry research company. One is the customer, voice of the customer. And two I always say is earnings. So we’re not going to talk about anything about that today, but I always say earnings is where you actually see if people are buying the product. And I say the customer’s the one that tells you if it’s any good. And by the way, the two things tend to be highly interconnected. Now, again, sometimes companies have great products and can’t execute, but most of the time when you have a great product and you’re getting great results, it’s because you’re a great company.

I’d love to talk a little bit about, this can come through the lens of you, the former CIO, this can also come from you, because I know you’re talking to many of your people that are now doing the role you were doing at Foundation Medicine, and that’s all about choosing technology. At the very root of what we do is infrastructure, meaning that in the end, our applications run because of the infrastructure choices that we make. You were very involved in modernization, you’ve been very involved in helping other companies. How do you position that whole selection? You went to, by the way, AWS, because you had actually, I believe you told me you brought your whole company up on it. So you’re a true believer, you put your money where your mouth is, but how do you position the choice and how do you talk to your customers about why it matters where their infrastructure lands?

Tom Godden: Well, there’s a lot happening today, and customers are being asked to, well, reduce costs, reduce risk, reduce complexity, while also needing to build capabilities to operate in a reliable, scalable, available manner. And oh, yeah, by the way, needing to increase speed, innovation, and value. And it’s a big challenge to be able to do that as an executive, and you need a partner who can help them do just that. And at AWS, we like to say there is no compression algorithm for experience. Doing this at scale longer than any other cloud provider, well, it matters. But it’s not just about our capabilities where we offer the deepest and broadest set of features, but it’s about our customer obsession to help customers meet their needs.

At AWS, the culture of innovation is driven by that obsession to help our customers, and it’s in our DNA, and it propels us to be able to help our customers meet these challenging needs. And when I was CIO, we had challenges along the way in being able to move the world’s largest genomics platform to the cloud. And AWS was our willing partner in that, helping us innovate and building new capabilities and solutions directly for that. And so finding a company that has the capability, has the skills, those are the table stakes, who has the broadest breadth and depth of features, capabilities, instances, regions, global. Yeah, all those things matter, but finding that customer who’s willing to be obsessed about your success, willing to innovate on your behalf, to me, that’s what was so intriguing as a customer and really brought me to wanting to work for AWS and to be part of it and to see what it was like on the other side. And I can tell you it’s even more impressive once you get on the inside than it was on the outside, and it was pretty darn impressive being a customer of AWS.

Daniel Newman: Yeah, you definitely were able to, it sounds like parlay the experience you had on as being a customer and now moving over to the vendor, and hopefully you’re probably mapping how you do that at scale. I mean, the hardest thing to do as a company that’s got the scale that you have is to give that deep level of service to every single customer. And of course, you do have good competition, so staying ahead of the competition is going to be something you’re going to focus on. And I think that’s something that has been very impressive to me about AWS. I noted a lot of really significant innovation that you announced at this year’s re:Invent, new chipsets, new services. Of course, I’m a salesperson. I love talking silicon, so I’ll give you that for instance though, I think 10 years ago now was the first inroads that were made by AWS, it’s a vertical integration, Nitro instances, first instances of Graviton, first instances of Inferentia, then Trainium, now Trainium 2, Inferentia 2. What are we on, the fourth generation of Graviton?

Tom Godden: Yep.

Daniel Newman: Nitro has continued to advance and the whole homegrown silicon and being able to optimize people’s environments and then do that all the way out to creating more performant, more cost-effective, more sustainable instances. But at the same time, you’re still very, I’ll call it Switzerland. You offer Intel, you offer AMD, you offer NVIDIA, you do try to make it flexible for your customers. But talk a little bit about all the investments. I actually, I have my philosophy, Tom. What story do you tell when asked about why you made those investments? Why in silicon so deep? Why in so many regions? Why so many services? Some could say that doing all that gives too much choice, for instance. Why is AWS so focused on making so many investments to meet these customer needs?

Tom Godden: Yeah, you mentioned that we’re aware of what our competitors are doing, and I would tell you it’s not where our focus lies. Our focus lies on understanding the problems that our customers are facing, the opportunities that our customers have, and then helping build solutions and working backwards from that to be able to innovate on their behalf and as we innovate on their behalf to be able to do these things. That’s what led us to build our own infrastructure. That’s what led us to design chips and processors to provide specific capabilities. And as you mentioned, to provide choice with our partnership with Intel and AMD and NVIDIA. With Intel, we just announced that at re:Invent, a new EC2 instance, powered by the custom fourth generation Intel Xeon scalable processors.

These custom processors available only on AWS offer a 15% better performance over comparable x86 processors used by other cloud providers. What’s important about that, Daniel, is we’re constantly looking at ways to do things more efficiently and effectively for our customers. Never have I seen a company that spends so much time trying to help their customers reduce cost on what we’re charging them by providing them more efficient, more capable things. And again, we’re doing that by identifying a problem that they had and working backwards to come up with that solution. And in doing that, you’re right, it’s led us to have lots of different choices across those instances.

Daniel Newman: Yeah, you’ve built, like I said, a ton of services. And by the way, you keep moving up the stack. If you look at the service progression, I’d love to see that on a dot plot, I think it would be a long piece of tape at this point, but this service progression and now really into everything serverless and containers and application, EDA, all the things you’re doing at one time, it’s a lot. It’s a lot to have all this expertise. And of course, you’re also moving more into applications themselves and kind of expectedly in many ways. But I think you’ve also been very targeted. You’re not rolling out word processing just yet or anything like that. You’re trying to focus on applications that make sense for your customers, and they can collaborate, connect, are ochestratable with their existing enterprise ecosystem. What about managed services, Tom? The company seems to be very focused on that. It seems to be focused on, say, serverless for instance, but it seems to be very TCO, time to value focused. What’s the thoughts there?

Tom Godden: Yeah, we pioneered serverless movement back in 2014 with the launch of AWS Lambda. Now I can’t imagine a world now without it. It seems like so obvious and so capable in what we’re doing. And now we have a full portfolio of serverless services across compute, storage, databases, analytics, event driven architectures, and even more, some of them, like DynamoDB, are fully managed, NoSQL database or S3 are object storage, EFS, elastic file storage, and Aurora. I love Aurora, our fully managed relational database. Customers love these innovations, these new services where AWS, well, we do the heavy lifting and manage the complex infrastructure. Will customers benefit from all the automatic scaling and the custom silicon we built this on? You mentioned Nitro, that built-in security and reliability while giving them the opportunity to have lower cost and faster time to market. I mean, it’s like getting your cake and eating it too, and these services are hugely popular.

To give you an idea, on Lambda, we have over one million active customers each month using Lambda processing, tens of trillions, let that sink in for a second, tens of trillions of requests per month. And now we just came out with another innovation working backwards from a need our customers had to launch Lambda SnapStart. SnapStart delivers up to 10 times faster startup performance for a Java function. So when you first launch it, 10 times faster. And the great thing is you can enable this and get these performance capabilities typically with no code changes, you just benefit from these innovations where we identified, again, a need that a customer had and we innovated on their behalf. It’s just exciting to see these things come out time and time again and hear customers go, “Oh, that’s exactly what I needed.”

Daniel Newman: So trick question time, is there really no servers in serverless?

Tom Godden: Yeah, yeah.

Daniel Newman: No answer required.

Tom Godden: Yeah, no answer required. Yeah, just go silent, fade to black.

Daniel Newman: Let’s just say that sometimes the particular hardware matters and sometimes just running the app matters, right?

Tom Godden: Yeah.

Daniel Newman: And I joke, but-

Tom Godden: Well, and in serious you joke, but here’s important about that, our track record of being able to do this at scale is where we’re able to say, yeah, there is a server, but we got it. We got this taken care of. Don’t worry about it.

Daniel Newman: It’s all about being truly a managed service. And when you make it managed, then the customer’s not worried about the instance, the silicon, the setup. As far as they’re concerned, it’s just the app. So let’s pivot for a minute here to gen AI. I’m really proud about that, by the way.

Tom Godden: Come on, I’ve never heard of gen AI before. Could you explain to me about what it is?

Daniel Newman: 14 minutes in a conversation about technology and we didn’t really say AI yet. I mean, I kind of did reference the chipsets, but other than that, not so bad. So here’s the thing, I’ll be candid because my job as an analyst is to read the market. Mid-year I was a bit concerned about AWS. I thought you guys got out the gate a little bit slow. The best thing though that happened to me is I went to New York, you had this big generative AI event for analysts. You brought us into a room. I kind of read through at the time what was your internal kind of viewpoint. It’s another Amazon thing. It was like a 14-page thing. We sat in a quiet room, read it together. You know that drill, right?

Tom Godden: Yep.

Daniel Newman: And I learned a lot there. And the one thing I learned is that just as methodical as you are with those 14 page or 10 or whatever, it was long, you were methodical about how you decided to come to market with generative AI. A lot of companies were kind of doing the uh-oh, better react, sprint to market. We saw some big fumbles on early iterations of what came out. You all waited, but I would say that as the year progressed, you’ve been, it’s a very tortoise and hare thing that happened. And while I’m not calling you a tortoise because you move quick, but very methodical, very thoughtful, very open. And I actually think within from summer to winter, you’ve made a remarkable stride in terms of generative AI, whether that’s, we talked Inferentia and Trainium, but we also, we can talk about SageMaker, you could talk about Q, you could talk about you have so many services to help companies. Talk a little bit about how you’re approaching generative AI and how you’re helping customers really think about it maybe methodically, proactively, and successfully in their business.

Tom Godden: So generative AI has unlocked so many new opportunities for us, and it’s really exciting and we believe we played a key role in democratizing AI and making it available to anyone who wants to use it, including over a hundred thousand customers that are using AI. So this idea that we’re a little bit late, we’ve been doing this for years with amazon.com and innovating and using artificial intelligence and machine learning for a long time. But when it comes to generative AI in particular, we like to think about it as three layers. And we’re investing in all three of those layers from infrastructure to services to applications. You mentioned on the bottom layer of that infrastructure is the silicon that’s used to train these foundation models and put them into production. But part of it is also our deep partnership with NVIDIA. And we also then have those Trainium and Inferentia chips that we talked about earlier. But as we move up that stack, you need the right tools to be able to train and deploy all of this. And we do that through our services layer, and that’s where we have Amazon SageMaker, our managed service that makes it easy for developers to build, train, tune, deploy, all those machine learning and foundation models. And now we’ve added Amazon Bedrock to that. And Bedrock is the easiest way we believe to scale and build generative AI applications with large language models. And what’s nice about Bedrock and different from what is out there in the industry, Bedrock gives you enterprise grade security and privacy because we designed it that way from day one. That’s in our ethos, that’s in who we are. But Bedrock also importantly provides you choice. It provides you choice of models. You see, Daniel, we don’t believe that there’s going to be one large language model to rule them all.

You’re going to need choice. And today, Bedrock provides a wide variety of those choices, whether it’s AI21 Labs, or Anthropic or Cohere or Stability AI or our own Amazon type model. But we also didn’t stop there. We also have the first to make available Meta’s Llama 2 model available through that fully managed, you know, API, we have over 10,000 customers that are using Bedrock, including Adidas, that’s enabling developers to get quick and accurate answers to technical questions and Carrier who’s helping people create more energy efficient homes. But then that brings us to that last top layer. We’ve built applications on top of foundation models so you can get as a customer, quick advantage and benefit from generative AI without needing to get into the specifics of it. And one of those is Amazon CodeWhisperer. And Amazon CodeWhisperer as a CIO is one of those few things that just blows me away. I’ve watched the productivity that people can get with CodeWhisperer, which uses generative AI to provide real-time coding suggestions within your development environment. We’ve seen people be able to get up to 57% faster time to value with using CodeWhisperer and 27% more likely to be successful with the code that they’re writing. I mean, oh my gosh. And now with the ability to add your own customizations in your own library to CodeWhisperer, and that’s an important differentiation to our generative AI approach. We want to bring these models, we want to bring these capabilities into your secure environment so your data remains your data. You’re not taking your data and pushing it out into these public large language models and training it for other people. And that’s what brings us now really exciting at re:Invent, we announced Amazon Q, a new type of generative AI powered assistant designed to work for you and to help you at work in how it operates. We’re really excited about those capabilities coming out.

Daniel Newman: Yeah, absolutely. And your enthusiasm, I love to see it. Look, I mean, something about CIOs and we work closely with a lot of practitioner and practitioner leaders. CIOs are really what I call the translators between the executive branch and the technical branch. And it’s a really tough role, because you’re talking generally, everyone that works from you knows more about tech because they’re the doers. I say that, but you know what I mean?

Tom Godden: No, but you’re right. You’re right. I know Excel and PowerPoint at this point in my career.

Daniel Newman: Everyone on the executive team knows more about the financials and the… So you’re trying to figure out how to be this conduit between these things. But when you become enthusiastic, it’s generally because it has so much value on both ends, technical value and business value at one time. And so seeing your enthusiasm really makes me smile.

Tom Godden: Awesome.

Daniel Newman: And by the way, I’m really glad you mentioned Bedrock because I mentioned a few of the services, but that was one that actually really caught my attention. I like the open approach. And by the way, other providers that kind of went down the, we’re going to black box it or we’re going to kind of own it, I don’t know about you, but I’m already seeing the pivot of like, oh, well maybe we’re not going to make everything run just on this one model that we control. Because ultimately customers are going to want choice, they’re going to want flexibility. And by the way, there’s probably never going to be one model that’s best at everything.

Tom Godden: Well, exactly. Because you have models that are better at conversation, you have models that are better at summarization, you have models that are better at creating obviously imagery and that type of thing. And that’s where we believe you need to be able to quickly do that. But here’s the other thing, I don’t think that the models have sorted themselves out yet on which ones are going to be the preeminent models yet. And if you invest deeply, embed and ingrain in one, that may not be the one that is, let’s call it the victor, the preeminent model. You’re going to struggle to make a pivot when something else emerges that is more defining. So I think it’s still early days in this, and that’s why we’re looking at this going, you need choice, you need security, you need to be able to manage privacy. And that’s why, as you said, deliberate approach sometimes has a connotation of being slow. I don’t think we were slow, but we were very deliberate.

Daniel Newman: Yeah. And by the way, I want to be fair, I wasn’t by any means suggesting, part of what has made the catch up so fast is the fact that you were already investing. The second thing is you’re running the most workloads on your infrastructure. And so yes, there is this sort of hype about these kind of AI data centers, but most people’s apps and data are sitting by far the largest provider of those services is AWS. So there is a fairly significant reason to believe that once you start, every AI service you layer on is going to be adopted with a pretty good scale. So that was a good thing. I just didn’t think the communication… So when I was saying that, I was just talking about the market was kind of waiting. Everybody was sitting back like, when are they coming? And that’s what I said, it was very methodical and thoughtful. I wasn’t so much, wasn’t worried. I was wondering as why are you quiet? Why are you not talking? But it’s making a lot of sense now. So let me put you back in the decision maker role, the role you had when you were moving workloads as a CIO. How do you frame the decision making process and helping builders take advantage, by the way, take advantage of all AWS? That’s one of the things I get stuck because I sometimes look at it, it’s like the Cheesecake Factory. There’s like 1,000 things on the menu. And my one idea was to just have Q do it, but I’m not sure we’re quite there yet. So how does a CIO, how does an enterprise maximize and get the right services and become as successful as you were with their AWS deployments?

Tom Godden: Well, one thing I would just tell people to start out with some of their, what was announced at Keynote. Go hit YouTube and watch Adam Selipsky and Werner Vogels and others talk about the things that we’ve done and what we’ve innovated. But I also think there’s an opportunity of people to dive into the developer communities and connect with the AWS heroes to really learn how their peers are building things out. We’ve also invested significantly in the amount of training that we’ve put out there to make it available, both free and paid training, various different certification programs. Something I wish I had done as a CIO was trained more of my teams faster than I did. When I did that is when I saw through transformation ultimately occur. And you can’t underestimate the value of a good partner to really accelerate and bring an understanding of all of these capabilities and best practices into this.

So I encourage people to look through that AWS partner network as they look through some of these solutions. And then candidly, a shameless plug, our team’s here to help. Our team is here to help share that experience and that knowledge around people, process, organization, the culture. Daniel, it turns out those carbon-based life forms matter. The technology’s important, but the people matter in this and we got to put them in the right position to capitalize on all this wonderful technology. When you combine the two of them, which is the reason why our role is so exciting, when you combine world-class technology with the right organizations to drive innovation and speed, man, customers can really do some exciting things. And that’s why it’s a lot of fun to be in this role and to help them out to do that.

Daniel Newman: Yeah, very exciting. Very exciting. Listen, Tom, it’s been great having you on the show. Any final words for the audience before I let you get back to your day?

Tom Godden: Well, just keep innovating. I’m impressed every day as I engage with customers and what they’re building and what they’re doing. And we’re excited to play a part and be customer obsessed and helping them get the right solutions to do that. So thank you very much for having me on. Excited to see where everyone takes us.

Daniel Newman: Yeah, Tom, love to have you back sometime soon. Thanks a lot.

Tom Godden: Thank you.

Daniel Newman: All right, everybody hit that subscribe button. Join us for all episodes of the Futurum Tech podcast here on The Futurum Group. For this episode, though it’s time to say goodbye. Thank you, Tom. Thank you AWS for making the time. We’ll see you all later.