Kim Majerus:
I'll send you along and you could grab some folks to go as well. I'm going to change a little bit. And you mentioned startups numerous times through the course of your keynote and the CIO and just now, and it's obvious, it's very near and dear to your heart. And I still like to think that AWS operates as the largest startup that's out there.
Matt Garman:
That's right.
Kim Majerus:
How are we helping... I'd like for you to share how are we helping those startups innovate on behalf and in partnership with the other organizations that are out here?
Matt Garman:
There's a couple of things we do. Number one is we provide a lot of support to startups. We said next year we'll fund more than a billion dollars in cloud credits, helping startups get off the ground, helping startups scale and grow. Look, it's not charity. These are the enterprises of tomorrow and we want them starting and building on AWS. We think we're the best platform and the best way for them to get started and for them to grow their businesses. And so we want to help. And actually, I was just at a VC round table this morning. Also, another important part of that whole ecosystem is getting... And we have a tight partnership with many of the venture capitalists as well as the startups. The other way though, and this is an area where I think AWS is uniquely helps startups, which is we've increasingly been spending a lot of our attention and focus on helping startups go to market and thinking about how they can reach a number of enterprises as well.
And I think that that matchmaking is incredibly important because there's a bunch of these startups that have really innovative solutions that could help many of the largest companies in the world, but just getting visibility, understanding how they work, integrating them in is something that they're often not scaled to do. And when we're able to make those connections, it's obviously great for the startups because they get customers and it's great for the enterprises because get exposure to incredible technologies or products that they otherwise wouldn't have. And so we've been spending a lot of time on getting that flywheel together and it's one of the things that I'm seeing more and more startups really value is that joint go-to-market effort where we can help them get connected in with the rest of the AWS ecosystem.
Kim Majerus:
I think it's the adjacencies and the opportunities for those established organizations. Do you have any recommendations for those leaders out here? How to get in touch or engage in that environment a little bit more effectively from your purview?
Matt Garman:
Well, I think that the partner organization does a fantastic job of making those connections. And for you all that are enterprises and established companies, your sales team can help connect you there, but marketplace is a fantastic place to find some of those and we make it easy to procure through the marketplace, but letting our teams know just what's interesting to you, the spaces that you're interested in, can help us try to do some of that matchmaking. And we've built some technology to help on that too so that we can do it at scale, but I think that's an area where we'd be delighted to help on that front. And then if you're a startup, the startup team has deep connections into the partner organization so you can get your products up on marketplace and start to be connected to customers that way.
Kim Majerus:
It's a great way and it's those proof of concepts and those opportunities, which then leads me back to generative AI. Over the past 18 months we've worked very closely with customers to help them understand the power of opportunity, what Gen AI can provide. What do you see from an enterprise perspective to actually get them to move from proof of concept to full production? And what are the commonalities in those enterprises or startups that you see?
Matt Garman:
Well, that is, I think the conversation has almost entirely shifted to that, which is we have a bunch of proof of concepts. We've done hundreds of maybe proof of concepts. And the interesting thing about when you do these proof of concepts is you didn't have to do a lot of the hard work to get those up and running. You did it quickly, you're able to get it up in your website or you get it into some internal app, but it's not necessarily plumbed into all of your enterprise data. It might not be plumbed into all of your permissioning and security and auditing because it wasn't part of your production systems. And now you have maybe 100 of these and you find the two or three or five or whatever that seem like they have high potential. The tricky part is it's not just, okay, let's do more of that one or keep that one.
You actually almost have to start over from scratch and think about how does it get integrated into your data? How do you really get value out of it? Because the key to getting value out of these applications is actually thinking about the ROI. It's not just throwing something out there and seeing what sticks. It is thinking about how do you drive costs down and how do you drive value up. It's not any different than the rest of the parts of your... It's how you think about which S3 skews to use. It's how you think about which database to use. It's how you think about which replication schemes to use, all of that kind of stuff you think about. And AI is a key part of that too. You want to think about where are you're going to deliver real value to your enterprise and to your customers.
And a lot of that is going to mean it's got to be plumbed into your sources of enterprise data. And so that data piece is where almost always where I see the most value coming back to is that it's deeply tied in to the sources of unique data that you all have. Because if it's not tied into your unique data, then it's just the same model that everybody else has access to. And maybe you have a unique way of using it, and maybe it's probably valuable, don't get me wrong, but it's not going to be the hugely differentiating thing if you're not bringing your own data. And so that is an incredibly important part of this, and it's where I actually see a lot of the focus going now to, okay, how do I make sure that all of my data is in a cloud environment?
How do I make sure that all of my data is labeled so that I understand where it is, what it is, that it's accessible, it's in a data lake that I can access and then in some sort of reg index or some way that I can then start to integrate that into the generative AI application where I think I'm going to get all that value. And so it's interesting because it's a lot of companies going back to the like, oh, I was going to do this whole migration and modernization thing. I paused it because I thought generative AI was going to be the new important thing. And I realized I still have to go do that first thing before I actually get most of the value out of the generative AI. You can do some of them in parallel by the way, so it doesn't have to be fully one or the other, but really getting all of that data into a great spot is incredibly important. And that's where I see a lot of that focus shifting back to.
Kim Majerus:
Right back down to the foundation of where it's at?
Matt Garman:
Yep.
Kim Majerus:
And I guess the question would be when you think about, when you talk to the customers and the stories that you get through the course of this re-invent activities that you've had this week, have you seen a key success factor in the organization or their approach that would be valuable to the audience today?
Matt Garman:
Well, I don't know if there's one, but I'm happy to go through a couple of them. One, on the data side, it definitely is if you can have your data in a queryable format in a single data lake, but also you need to have some organization and information about that. If you just have blobs of data without information or organization around that data, it actually is less useful. And so that metadata about your objects, the metadata about your enterprise data is actually super useful and important that you arrange that in the right way so that you can actually get value out of it. Otherwise, your AI systems, they're not going to be able to just understand the blobs of data without context to them or you're not going to know where to go after, so that's a super important part. I think we can help and we're starting to build capabilities for you to go add that data to your objects, your data that you don't have that. But that's an important one, and that's where I see a lot of success.
I do think that where the technology is right now, there is a huge opportunity to drive efficiencies in the organization with things like Q. It is, we are seeing leaps and bounds improvements into how effective those conversational assistants are to everyday work life. And if you're a developer, I think there is massive opportunity on the development side where we're talking maybe two, three, 4X the time, the efficiency that you could get out of a developer. In fact, out my keynote yesterday, I have a friend who's a developer at Amazon and he sent me a text and his text to me was literally, "I wish I had coded for an hour a day." It's actually less than an hour a day that he usually gets to code. And so it's actually, it's both not surprising and shocking at the same time, that is less, that's how much time that a developer or senior developer codes. And so if you can take an hour a day and make it three hours a day, it seems like that's a low bar, but you just gave 3X of productivity of what you're getting out of your development teams.
And development teams are not cheap. There's not developers falling off of trees. They're a scarce resource that are expensive. And if you can get two to 3X the value out of when they're delivering for you all, that is an enormous gain. Think about how much more you could do if you had three times the size of the development team, building generative AI applications for you, building the next generation experiences for your customer. It is a huge lever and I think that kind of savings is doable across many different roles. We see that huge wins in the contact center space. Connect has added a ton of generative AI capabilities, and if you look, most of our Connect launches happened over the last couple of weeks, but there's some real material wins there both in handling things like email, handling things like chat. It turns out that the customers actually like it better. It's not just cheaper and easier and faster for you, it's easier for the customers too. And so your CSAT will go way up by having these generative AI experiences where customers can quickly get the answer that they need because you don't have to wait on the phone, you don't have to wait for a person to bumble around for an answer, you can actually get it really quickly and move on. And as the systems get more and more information about your customers, they get better and better about answering those questions. So those are a couple of places where we've seen massive gains and where I encourage you all to lean in and see where you can build there.
Kim Majerus:
We have several use cases. Connect is a great example.
Matt Garman:
I know Connect, is one of your favorites.
Kim Majerus:
It's my favorite actually. PQ, if you're out there, thank you for all the great work you do there.
When I think about the citizen engagement and what we're able to achieve to ensure that people are getting the right information at the right time, effectively, not sitting on hold for hours or pausing. It has made a transformational experience for our citizens. And that's the beauty about being in public sector, you could see the immediate impact.
And when I think about GenAI, there's so many great programs that are available and Francesca talked yesterday at the CIO Summit or the Council about the work that the GenAI Cloud Innovation Center is doing. So if you have GenAI opportunities that you want to work with AWS, obviously we've got a team that is more than happy to jump in. Just a little plug from the Pub Sec Summit that we had in DC, we also did a two-year $50 million investment designed to help public sector organizations accelerate innovation and support their critical missions. It's a great opportunity to have those who are interested in the public sector space to embrace and engage with your AWS counterparts to help you get there as well.
Matt Garman:
It's a good point. Just one more to add to that, Kim, is the GenAI Innovation Team, if you all haven't taken advantage of that team I recommend reaching out to them because one of the things they're actually quite good at is that thing that I mentioned before, which is sparking that creativity of how you can really use this technology to transform your business. And they'll sit down with you for a half-day session or a day session and help bring ideas of what's even possible with the technology, what the technology is capable of, what they've seen it be successful in your particular industry. And that can be a really powerful tool if you haven't all used them. And so I encourage you to reach out if you haven't.
Kim Majerus:
Our customers love it and they want to continue to engage and grab more out of it. So we'll definitely make sure that we get that information out there.
Back to how we think about startups and the organizations and really seeing true value from the startup community or through those established organizations, what do you see as the needle mover from a GenAI with the startup organizations?
Matt Garman:
With startups in particular?
Kim Majerus:
Yeah, particularly.
Matt Garman:
I think that one of the benefits that startups have, I think that they don't have legacy or preconceived notions of exactly how things should work. And so, the hard part of re-imagining how your business should work is you know what works or has worked in the past. And so you have all of this kind of preconceived set of notions of, "Well, obviously I have to do this this way." And I think startup founders are, at least the good ones, are really good about seeing how something could be done differently. And so, they're very good at taking technology and looking around and saying, "If we apply it in a particular way," and they're always willing to really lean into new technologies too, which I think is fantastic. If there's a new model that comes out or there's a new framework that comes out, startups, like that they'll flip from the old one to the new one and they're going to try the latest thing that can help them move fast.
But that's why it's fun and exciting to work with the startups. And I think that's where the ones that shift quickly and really adapt those new technologies are often the ones that go really fast and do grow significantly. Because when you're a startup, the trickiest part is finding that product market fit, and as soon as you get that, and so you've got to kind of shift a little bit until you get that. But hopefully you saw from a couple of the videos yesterday, there's some people doing some fascinating things. These are startups, if you think about some of those, that really could completely change how manufacturing works or how software is written or how medicine is made. Those are incredibly ambitious and big things to change. It's pretty fun to see that where, as an incumbent or an established company, you don't as much lean in.
And I actually think one of the cool things that we try to encourage our teams to do is to think like startups. And I would encourage you all to push your teams a little bit to think like startups and push teams to figure out how they can go faster, how they can adopt new technologies more quickly. And think about mechanistically. It's different, you may not be in a... Assuming you're not a startup now, if you're in a larger company, you do have constraints, you have audits and requirements and other things like that. But it's useful to think about what is a mechanism where you can give your teams permission to experiment. You can give your teams permission to try new technologies. The cloud is a great place to do that. It's easy. You don't have to go buy a bunch of servers, you don't have to build a data center. You can experiment with stuff.
And so telling your teams, "Look, go experiment with Q Business and see if we can find some interesting things. Go experiment with Aurora DSQL and just play around with it and see if there's something interesting that you can come up with. Go play around with the new S3 metadata and think differently about where we might be able to use that." Thinking mechanistically about how you get your teams out of there, "This is how I've done it for 10 years or 20 years or 50 years, and so I'm always going to do it this way," is super important. And I think that is a lot of times that's the difference between enterprises that are on a flat to slow decline to ones that continue to grow. It's hard to do, it takes intentional work when you have a large established company, but super important.
Kim Majerus:
I guess re:Invent is the best place to start that.
Matt Garman:
That's right.
Kim Majerus:
I'm going to ask you a question. Yesterday we heard during Andy's fireside chat, where is Andy's spending his time, where he thought he was going to spend versus where he is? I guess that would be my question too. Is there a specific industry that you find yourself paying a little bit more attention to because you see the opportunity for innovation and change?
Matt Garman:
Well, there's two things that I'll say. One, from an outside perspective, I don't know if it's surprising or not but I do actually spend a good amount of time with our government customers because I think that there's an enormous potential in government to go... I mean, anyone thinks our government moves fast, maybe you do, but not many people think that the government is agile and at the leading edge of technology. But I think they can be and there's many folks inside of the government that are. And that's the actually cool part is that there's really talented forward-leaning folks inside of the government that are working hard to make changes, and it's fun to find those people and to really lean in and help them invent because I think we can together kind of power them to actually make really big changes across countries and across citizen groups and things like that.
So that has been fun to spend a bit more time at. And getting to meet with some of the world leaders and talk about that has been cool and an interesting and new experience for me that I get to do. I think, it's not exactly what you asked, but also one of the ways that I've been also spending time thinking about is exactly what I was just suggesting that you all do, which is think about mechanisms for our own internal AWS. How do we continue to make sure that we can invent at the speed that we are? And it's a large organization and we do move really fast, but it's still, even if things are going well, you can't be complacent about that.
I have spent a bunch of time just learning how teams are organized, learning where there are early signs of inefficiency or slowness or where things could be better, and thinking about how we identify mechanisms for that, how we push decision-making down so that people can go faster in the organization and not feel like they're burdened with having to have a bunch of people approve things before anything can get done.
And I think there's opportunities for us to do even more of that, but that's where I spend my time because I actually think that's where I can have leverage sometimes. I like to think... For those of you don't know, my first job at Amazon I was the product manager for AWS, so I was the only product manager for all of our AWS Services. And sometimes I like to think about myself still as product manager, but I have to remind myself that's maybe not the most leveraged place to do it, even though I enjoy doing it sometimes.
But anyway, so that's where I get to spend some time. But I also then sometimes cheat and actually do spend time in reviewing new product ideas as well because I like to do that as well.
Kim Majerus:
Well, we do know you like to dive deep, that is high on your list of superpowers. So it's exciting too. And thank you for spending the time with the public sector customers. We always enjoy having you provide them those innovation opportunities because you're absolutely right there's great innovators within the constraints of government and they just want to do mission work, and that's exciting.
Matt Garman:
That's right.