Dan Schlatter/Business Journal

Andrew Jassy is senior vice president of web services at Amazon.com. He heads the company’s “cloud computing” business, which uses the vast infrastructure Amazon has built over the years to run its ecommerce engine, and makes some of that capacity available to web developers in the form of data storage, computing power, and other services.

I sat down recently with Jassy to talk about Amazon's web services business and how much progress it's made attracting large enterprise customers. I also asked him about future product releases and the recent hiring of a data-center expert from Microsoft. Jassy filled me in on some of his extracurricular activities as well, including competitive wing-eating and watching football in his sports-bar style basement.

Q: Amazon Web Services has had a lot of web startups as early adopters. How important is it to have large enterprise customers on board, and what kind of progress are you making on that front? Are large enterprise customers using Amazon Web Services for experimentation or building core services and products?

Jassy: We had always anticipated that our services would initially appeal to the startup space and that's because with any big computing shift like web services or cloud computing, it's often the smaller customers that get the revolution going. They see the value proposition soonest. They have very little capital themselves and it really levels the playing field, which is particularly attractive to the startup space. What we figured would happen is some larger companies, some pioneering ones, would kick the tires a little bit.

But mostly they would take a few years to kind of see the services get launched, see what they were, see if they actually perform operationally, and see if they're getting adopted. And part of what's happened that's surprised us, is that the enterprise interest in adoption is happening much more quickly than we anticipated.

We actually see good progress in the enterprise segment. We obviously have a very significant set of startups using the services, but we have enterprise customers like Eli Lilly and Autodesk and Nasdaq and New York Times and the Indianapolis Motor Speedway, and a host of others I can't mention just yet, in addition to some pretty significant enterprise partnerships with Oracle, with Red Hat, with Sun, with MySQL, with Facebook, and Saleforce.com. And that is actually happening much sooner than we anticipated, which makes sense to us because the value proposition for big companies is very compelling as well.

They spend more total capital on this infrastructure piece of their technology than the smaller companies and especially when you look at this economic climate, if there's a way to turn Cap Ex (capital expenditure) into Op Ex (operating expenditure) and do it more quickly, it's a huge advantage, as long as who you're working with is reliable and scalable and cost-effective. And in terms of your other question on usage, we see people doing everything from proofs of concepts to pilots to actually running core parts of their applications on web services. It's all moving at warp speed right now.

Q: Can you give an example of companies that are using web services for a mission-critical application?

Jassy:  I don't know what we can disclose or not disclose on that.  

Q: I was trying to get at the distinction between dabbling with it and actually building a critical piece of infrastructure with it.

Jassy: I'll give you a range of examples. There are some companies that are using this as a way for their sales forces to get out and do proofs of concepts with prospective customers. And it allows them to sell their applications much more quickly. They're able to provision proofs of concepts with their customers in minutes instead of weeks, so it gets them through the sales cycle quickly.

If you take Nasdaq as an example, they’re using our services to build a tool that makes it much easier for financial institutions to look at what historical prices have been for various stocks over time, so they can look at trends, they can make smart financial decisions on new entities they're thinking of investing in. If you look at the New York Times, and the New York Times had this idea, and they had it internally for ten years. They have archives of articles that go back to the late 1800s, I believe. They wanted to have an easy way for people to find them on the web, search for them and go buy them. With a couple developers they went and used our services to first convert all the files into digital files and then to process them in multiple sizes and put them up on the website and use our payments service to let people buy them.

They now have a service that has all New York Times articles from the late 1800s that people can find and buy. That was done in a matter of 48 to 72 hours, which is something they've been trying to find the budget and time to do for 10 years.  

Q: A broad strategic question: How committed is Amazon to web services as a business? Is Amazon going to put sales and technical support teams in places where customers are located? Is that part of the plan?

Jassy: We got asked this question I would say a lot more in the first year or two that we were doing it than we probably do now. But we still get asked the question: Is this an experiment for us, how committed are we to the business? And I can tell you this is in no uncertain terms a long-term commitment for us. We believe it’s going to be a very significant free cash flow generating business that’s going to be very meaningful over the long term, and that’s the feeling from the most senior levels up in the company all the way down.

And if you look at how the business has run and how it’s progressed over the last several years and what we’ve launched, it’s reflective of that. I think the operational performance is quite good. I think if you look at the speed of innovation, if you look in our compute service EC2 alone in the past year, we've launched static IPs, persistent storage, CPU heavy instances, Windows instances, we've launched in Europe, the service is in general availability with a Service Level Agreement, so we are very much committed to this business and extremely excited at the pace with which it's growing and view it as a long term investment and business for Amazon.  

Q: You mentioned SLAs (Service Level Agreements). Can you talk about the security/reliability question? I know there have been some outages over the past year. Where do things stand now and do Web Services plan to do more to increase security and reliability?

Jassy: For us, reliability and security are always our first priority and I think one reason that the services and the business have grown this fast in part is that the security and reliability have been quite good. But I can tell you whenever we have any kind of operational performance issue, we drop whatever we're working on no matter what it is and the entire team focuses on making sure that we've fixed the problem and we communicate appropriately. As I said, our operating performance has been quite good thus far, but we won't be satisfied until we're statistically indistinguishable from perfect.  

Q: How do you get there?

Jassy: I don't think there's any magic bullet on it. I think operating these services 24 hours a day, 7 days a week, 365 days a year at the performance levels that our customers expect -- startups and smaller companies expect it to be perfect, but obviously enterprises do as well -- it's not easy to do. It's going to be interesting over time as more people enter this space. We had to get good at this at Amazon to operate Amazon.com and one of the reasons we got in the (web services) business a few years ago was we realized we had a competency in running very reliable, scalable services at low cost to run Amazon.com. We thought we were pretty good at it 3 years ago. And I can tell you 3 years later, from operating these services under any imaginable use case -- and without people warning you that they're going to start using you at whatever volume they are. They don't need to warn you. They just start sending service requests to the API -- it's not easy to do it well and we're much better today at it than we were 3 years ago. And we thought we were pretty good 3 years ago. There really is no substitute for experience and for continuing to evaluate where you can improve and applying resource and effort at improving that.  

Q: Can you give us any peeks at what is to come in Amazon Web Services? What areas are you targeting for future releases?

Jassy: We just released about a couple weeks ago our first version of the AWS management console, which is really a web based GUI to provision and manage your AWS usage. And we launched it just for EC2 -- that was the first version of it. And we will launch it for all of our services. That's one thing you can expect: tools that will make it increasingly easier and less time-consuming for developers and businesses to use our services.

We are also working on monitoring and load balancing and auto-scaling services, which we expect to have out in the first half of this year. Those are things that we have talked about various times publicly. I can tell you we're working on several others services and we will continue to expand geographically.  

Q: How many people work in web services?

Jassy: We don't disclose how many people are in the business, but I will tell you we have a significant, meaningful, talented team working on this. And if you look at what's happened over the last year or two in terms of what the team has delivered, you can tell it's not a team of 10 people.  

Q: Would Amazon ever make acquisitions to grow AWS?

Jassy: We've made investments as a company. But web services for Amazon is no different from the retail and seller businesses that we also have which we're always looking for ways to grow the business and add capacity that are unique and that we don't think we can easily build ourselves or that we want to have quicker. In the web services space as well, if the right opportunity presented itself, we would consider it.   

Q: James Hamilton from Microsoft came over to your shop recently and he's done a lot of work with the concept of data centers made up of shipping containers.

Jassy: We have had to spend a lot of time over the years learning how to run very scalable, low-cost data centers given that we've been running a retail business, which is a very low-margin business. We're very thrilled to have James join the team. He's a very talented individual and has been spending a lot of time thinking about this space over the years. There are a number of areas where James is going to help us evolve our data center strategy, which is obviously important in this space. He's been here a few weeks, which makes him a veteran of sorts, but he will surely play a hand in helping us think this through over the long term.   

Q: Will Amazon be looking more seriously at the containerized model?

Jassy:  We're going to look at a number of different — there a lot of different ways you can go with data centers and that will be one of several approaches we will look at.   

Q: Does Amazon use containerized data centers now?

Jassy: We don't disclose that.   

Q: Now for a fun question: I understand you organize an event for friends and colleagues called the Tatanka Wing Eating Contest?

Jassy: That’s pretty good recon. This started 11 and a half years ago. We have an event we call Tatanka. It came from the movie “Dances with Wolves." It was the Indian name for “buffalo.” We get together the first Tuesday of every month at the Wingdome and we eat buffalo wings together. There are different membership levels you can choose to try and qualify for in Tatanka. It’s a lot of fun. It’s a great way to get people together outside of work, under a very important common theme, which is eating wings.  

Q: How do you rank?

Jassy: I’m a platinum member and a veggie member. To become a platinum member, in one session you have to have 25 wings …   

Q: Oh my God.

Jassy: … and then five pasty wings. And a pasty wing -- if you’ve ever had wings on a big tray, the sauce, if it sits at room temperature, it congeals and gets pasty -- so the pasty wing is a wing wrapped in that paste. A platinum is 25 wings plus five pasties. And veggie is 100 celery sticks plus 25 pasty celery sticks.   

Q: In one sitting?

Jassy: In one sitting.   

Q: And you did that?

Jassy: I didn’t do both veggie and platinum in one night. I did them in separate nights.   

Q: I read that you have 7 TVs in your basement.

Jassy: I am a sports fanatic, that's true. I have a sports bar I built in my basement called the “Helmet Head.” I have 7 TVs down there. I grew up in New York and my Dad, who's really a sports freak, had season tickets to the Rangers and Giants, so I went to those games very regularly as a kid. And when I moved out to Seattle, my friends and I were going out to sports bars on Sunday to watch football games and my dream was to be able to recreate that environment in my basement. So these same guys come to my basement now instead of going to the sports bars.   

Q: Do you retain your New York sensibility or have you become completely Seattle-ized?

Jassy:  As a sports fan, I have very much retained all my New York sports allegiances. I would say that I love it out here. I never expected to be out here long-term. I always expected I would be back in New York. But it's really hard to beat the quality of life and the beauty of it and the outdoor activities. And I'm really lucky to get to work at a place that is as exciting and engaging as Amazon is. I would say that people know I'm probably not from Seattle. I probably move at a little bit quicker pace than the average Seattle person. I probably drive a little more quickly and honk my horn a little more frequently. So I probably still have some New York in me. I still walk across the street against the red “Don't Walk” sign too.