Well, hello everybody. Here we are in lovely Las Vegas at IBM Tivoli Pulse 2009. And we've got a special guest with ourselves. Would you like to introduce yourself? Yes, Kotick. Nice to spend the time with you today. I'm Lauren States and I'm the Vice President of Cloud Computing, working in Tivoli. And the job that I do is to go out and work with clients and listen to what they're trying to do around cloud computing and use it to work with the strategy and development teams to inform our investment decisions. But the fun part of it is to get the environment stood up and see, well, how does it work? And what do we have to fix? And where are gaps? And where are they trying to push it? And it's really done a lot to help us accelerate our time to market. Since we're at Pulse, can you give us a summary of what the cloud-related announcements were that you guys had here? Yeah, we made a number of announcements today across our consulting services and software business. And I'll start with the Tivoli software business because that has helped us also develop some of our services offerings. Today we announced the Service Management Center for Cloud Computing, which is our umbrella offering, which we will put all of our cloud capabilities so that clients will know as they're deploying cloud what are the products that are most appropriate for the environment. And today we also announced Tivoli Service Automation Manager, which is kind of our integration framework for all of our different capabilities within the portfolio. What the Service Automation Manager does is something really simple. It allows the elastic provisioning of resources with the self-service component and then allows our clients to automate their workflow and also manage the lifecycle of the services so that things can come to end of life or end of service. And the work that we've done with clients as we've worked on rapid or elastic provisioning, there have been requirements to, you know, for instance, provide more consulting
2:04so that they can understand where they want to get started because I'm sure this happens with you. People talk about cloud computing, and then you have a great discussion, and then you sit there and say, "Well, what should we do next?" So we've created, and my team has done visioning workshops with our clients, and we've created this, announced a formal consulting services offering around cloud strategy and planning out of our GBS organization. And is the consulting sort of before and after, if you will? Because you were alluding to something interesting. So now that we have a cloud, what do we do with it? I mean, it sounds like there would be two opportunities there. Yeah, yeah, yeah. Actually, that's a great comment because, you know, for sure before, particularly now when you need to spend your money well, but, you know, we've discovered there's a lot of learning that goes into deploying a cloud. And there are also issues around organization, culture, governance, like with any system, really. Right. You know, and so, you know, that has to be addressed. And so there's a lot of opportunity for planning and strategy and integration. Yeah, that's an interesting rabbit hole to jump down there is the, I've always been a big fan of Conway's Law. That's like the system that an organization built pretty much matches one-to-one the communication paths in the organization that built it. And like you're saying, there's cultural and organizational changes. And I wonder in the cloud implementations you've seen, if there's sort of a Conway's Law thing that works maybe bi-directional or something. Like what are the organizational things that cloud computing sort of changes for you or drives? Well, so for instance, you know, if they build it, if you build it, will they come? Right. Right? And one of the very popular workloads that clients like to start with is application development and test because it's lumpy by nature and there's great efficiency to be had within the infrastructure. But then application developers, some clients have told me, like their own physical boxes.
4:05Right. Right? And they want the same one for development, you know, for test, for production. And so just because you build a development cloud, particularly in some organizations that provide shared services, then the lines of business are not obligated to use the shared services. Right. Yeah, I always think of those as the under-the-desk data centers. Under-the-desk data centers. You know, then there are issues of culture. I had a conversation with a client who, in this particular organization, they had a, I would call it like an ad tech team. It's more formal than that, but they introduced new technology through a formal organization. And tremendous proof of concept, great results, wanted to go on to the next stage. And we got into a conversation about operations, right? Because operations would have to change their processes. And, and... Because those would be the people who are sort of handed over this new... Right. ...this new cloud to take care of and make sure it's up and running all the time. Right. The people factor comes in, right? I mean, everything comes down to people. I've been in technology a long time. And the people factor... Right. ...is pretty constant. We get better gadgets and more fun things, but people, people remain people. They pretty much stay the same, right? Yeah, they pretty much stay the same. They don't, they don't shrink like every other gadget. Well, as we've done these cloud engagements, we have also started to have more conversations about acquisitions of services off-premise. And, you know, this idea that could I spill over my peak workload into an IBM server farm? You know, I've worked with IBM to do my testing. I've done it more on a static basis. I'd like to do it on demand. Right. And so that helped inform the second announcement that we made in services, which is test cloud services. So that companies can spill peak load over into an IBM server farm on demand and then release
6:05it. And as the name implies, is that for testing in QA or what, is it for anything? Well, you know, I think it'll start with testing in QA, but I can see a lot of applications beyond testing in QA like, you know, at the end of a quarter banks do analytics around fraud or, you know, perhaps as an airplane is being designed towards the end of the manufacturing process, there are more analytics that are done and it causes spikes in workload. So I think it's, it's, it's of some comfort to do it in an application test where there's not as much risk. But I think that it'll quickly flip over into. Right. It's, it sounds like a similar introduction that's virtualization had on some fronts where it kind of came into the developer and the testing organization. Sure. And now it's invasive. One thing that's always interesting to talk with cloud people about is, is that decision process that people are making between a public and a private cloud to use one of the nomenclatures. And what's the decision process that you see people go through if they want to physically own their own cloud, if you will, versus when they're willing to spill over into something else that they don't own so much? Yeah. I, at the south, I mean, that discussion comes at many angles. So there's some industries, organization, you know, by regulation cannot make their data public. So they can't use a public cloud. Or, you know, if you work with military or defense, typically they're not going to make their data public. So, I mean, there are issues of industry and regulation and data privacy. Certainly. The other conversation that I think is interesting, and in fact more interesting, is the discussion around business and value. And how can I deliver value to my client, my customer, my clients, my ecosystem, my suppliers, in the most effective fashion, right? And so do I want to source it internally through traditional IT?
8:07Or maybe I would source it through a public or a private vertical cloud. And so increasingly I've heard those discussions from line of business executives that they're making the decisions based on business, not based on technology. And I think that that's a subtle shift in the conversation because the early conversations were so technology focused because we're all in a discovery, definition, precision mode. But, you know, now it's real business and people are under pressure. And if there are ways to bring value into their organizations more quickly, like we had the IT director of the USGA here with us today talking about using our business resilience services for their organization so that they could bring online that capability. We have a small staff and it was cost-effective, but it was a business decision. So the sort of functionality that you might get with the public cloud might drive considering putting something out in the public, so to speak, that otherwise you would want to keep on premise. But, I mean, that sounds like what you're getting at with the sort of business discussion is there's sort of that risk management thing that goes on. Yeah, yeah. Like ultimately the value I'm going to get is worth whatever the risk may be. Whatever the risk, right? Or maybe there are more skills out there or maybe there are some things that lend themselves to public clouds like collaboration, for instance, lends itself. Social networking and that's why our Lotus division has been so aggressive with Lotus Live and hosted notes and the things that they've announced. Right. That's a good transition because that was the other thing I was interested in figuring out is what kind of applications you see people running on the cloud. I mean, here at Pulse, obviously, there's a lot of focus on the operations layer, which is fine and fantastic for us nerdy ops people, I guess. But at the end of the day, there has to be sort of an application running on the cloud, I guess, unless you're doing job sort of batch sort of things. And so what, I mean, what do you see people using or deploying on the cloud? What are they doing with it? I also live in that kind of gorpi service management world and so my experience with that has been more limited, but absolutely for sure things like collaboration and social networking out on the cloud.
10:22I think there are emerging things within industry verticals like learning management systems, collaboration across the educational ecosystem of K through 12 through universities. Yeah. We've got some work going on with Marist College. And I think that these verticals – in fact, I think the verticals will – this is my opinion – the verticals will drive what applications run on the cloud. Maybe more – there will be horizontal things like collaboration, social networking. Right, right. But I think the industry verticals, citizen services, you know, what could you do with health? What can you do in education? So it's almost as if the – like what you're saying is that the – instead of having to write generalized software that you can sort of run and then customize anywhere, there's a chance to have much more industry-specific customized stuff that you would have that you can deploy on the cloud. Yeah. That you can deploy on the cloud. Or a cloud, I should say. Right. On a cloud. On the cloud. The federated big cloud. Some cloud. You made a great comment about simplification or standardization. And that's another thing that we're seeing, that standardization also drives adoption of the cloud. If in an organization everything has to be customized and every piece of middleware is used because in a situation it was best of breed, so you end up with everything of everything. Right. It really decreases agility. And so the cloud also drives a movement for simplification, standardization, and maybe in a sense in some cases just good enough. Right, right, right. It's like always a nice opportunity if you can get in sort of at the ground floor, if you will, of a new technology. Is it sort of – you don't get locked into the old sort of cost structures and complexity structures of old things. And so you can hopefully make sure it doesn't explode out in complexity and cost. Yeah, but we don't have a good record in the IT industry, do we? We've always got to fund the smaller gadget. Right, because we have a hidden agenda.
12:23That's true, that's true. We've got to get that gorpy stuff sorted out. Another thing, a couple of the demos that you were giving and things that I've seen were based around submitting jobs to something and doing sort of batch job processing, which for the MQ people in the world and everything is a perfectly normal way to think. But I'm curious how widely used people are doing sort of batch job management or job things with the clouds that you see customers using. And then also, I've been thinking, you know, it seems like regular programmers, regular programmers, mainstream programmers, if you will, sort of don't think in a job-oriented nature. It's more of an ongoing process. And I wonder if there's – if you've seen people struggling about that or if it's just a crazy theory in my head. I mean, so the workload demo that we showed you this afternoon, which showed managing compute resources from an external provider, was not necessarily meant only to be batch. It was meant to show that we could do it and we could manage it within one service management operation. Right. And we could provide quality of service to the user. And that was exciting. We did that in our Rome lab. And, you know, as you saw, we've acquired Amazon Compute Resource. Right, right. Yeah, that was great to just throw some Amazon stuff into the pool. Yeah. And we maintain the service level. Right. And assured quality of service. You know, I think that this will continue to develop. And it'll have to – I mean, of course, evolve past batch computing, which is probably still 50% of all computing. The use case that we've done internally with our – we have a program called the Technology Adoption Program, which provides resources, compute, middleware to our developers, our technical community. Right. We have about 200,000 technical professionals at IBM and some of them like to do projects on their own and we provide them the resources.
14:26And some of these projects turn into products and services, which is great. When they took their environment, virtualized it, automated it, put in the self-service, they brought their costs down 83%. And they did it in less than 90 days. Right. So, I mean, that's the other great conversation. That's the speed angle or agility, as it were. Right. And that's a great conversation to have with executives because, particularly in these times, everybody is interested in taking out cost while still providing value and innovation and business agility. Right. And cloud, I believe, gives you both. You take out cost and you can improve agility and drive innovation. Well, on that note, thanks for taking the time to talk with us. Oh, it's a pleasure. Absolutely. - Thank you, thank you. - Yeah, well, we'll see you around.