How IBM’s Watson could change your business

Watson Group Chief Mike Rhodin talks about how IBM’s cognitive computing tech is rolling out to enterprises, developers and end users.

1 2 Page 2
Page 2 of 2

If it’s not good, then we try to figure out why is there conflicting information in the underlying knowledge base. Usually it has something to do with the way we’ve constructed the knowledge base underneath. But you have to be careful not to train the system to answer the questions you know because then you’re just writing a rules engine.

You always have a blind set of questions that no one has seen and those run against the knowledge base and they let you know how accurate you are against the blind set. Then you know whether the knowledge base is trained correctly. So you go through these training cycles to build up your knowledge base. Once the knowledge base is built, the APIs are clear, it’s very easy to build the applications at that point.

CIO: Our readers, for the most part, are IT executives. If they have ideas about Watson, how Watson might help their businesses, talk about the process. Who do they reach out to? Is it their traditional IBM contacts? Is it specifically this group?

Rhodin: Either works. If they reach out to their normal IBM contacts, they’re going to contact my group. Because it’s new and it’s new technology, I have a team of experts that engage with the clients to help them work through possible use cases.

A typical engagement might look like this: A customer goes to a session at a conference, hears about this, gets interested, they reach out. We have people that sit down and talk with them about things and they get an idea of what they might want to do. If there’s something really interesting there we’ll bring them here to Astor [Place]. We have what we call a Watson Experience Center where you actually see how Watson works in different professions and industries.

We picked examples that everybody can relate to. It will trigger the imagination to make them think about how it could be used in their industry. We use context like a doctor’s office because everybody’s been to a doctor’s office. If you could see how the technology might be used in a doctor’s office, you can then draw analogies on how it might be used in your environment. It takes about an hour to go through this set of experiences and then we have brainstorming rooms where we sit down with experts from your company and ours and start to narrow in on a use case and how that use case might operate.

Typically, we would work with the customer jointly on business cases, business models, return on investment (ROI) type scenarios around the technology, because we recognize it is new. It’s not a standard ‘send an RFP off to procurement and go get the best price on something,’ because there isn’t anything like it.

Instead, it’s how do I build a business case so I can justify this when I go home? We work with them on those models. We’re working now with hundreds of clients and partners, so we’re at the point where we’ve got sets of models that we can start from that help us tailor to [different] environments. These engagements are getting easier as we go forward.

CIO: I was going to ask you about the practical nature of that, given that probably everyone thinks, after hearing about Watson, that there’s some application to their industry. How do you avoid being overwhelmed by either not particularly applicable use cases or people who really aren’t ready to do this and are just kicking the intellectual tires here?

Rhodin: Typically, that’s a good problem to have, and it is a problem we have. Since we’ve opened this building we’ve had two to three clients in for half-day plus experiences every day in the center downstairs. We do go through iterative process with some of them.

CIO: You have centers around the world as well?

Rhodin: We do. But there’s only one of what you’re going to see downstairs. We will open satellites of it next year in Asia and Europe but there’s only one of these [Experience Centers] in its current form. It’s not a slide show. It’s very different.

The idea here is that we’re going to identify the use case that’s going to have a good business case and then we’ll actually work with the client on developing that into an application and then go on to production.

The business model is actually pretty straightforward. There’s some upfront consulting on coming up with the ROI in the business, a very low cost kind of thing, but it’s just working together on something and then when we select it we go through an implementation phase, which is like a services implementation phase, and then when you go into production. It’s a cloud-based, Software-as-a-Service business model.

There’s not one price for Watson. It really has to do with what the application is and the complexity of the application. We decide what the pricing model for the end state is going to be in the upfront state, in that consulting piece upfront that we call a cognitive value assessment. In that cognitive value assessment we actually decide what the business case is going to be and what pricing model makes sense. Is it per transaction, per user, per case kind of a business model? What makes sense for that particular use case?

CIO: When is Watson not applicable? What are you learning about where this is not the right answer?

Rhodin: Right now the important thing is figuring out if you know what information you want to put in the system, the curation process. If you say: Well, I want the Internet in it, we’re pretty sure that’s not the right use case. It’s just too big right now. There are ideas about how to get there someday, but that’s not the state of the technology today.

It’s really about coming up with the domain that you’re interested in and where you might want to go with that domain. That’s one of the key first indicators. The second is making sure about sponsorship of the project. Because it’s emerging technology, you need pretty high-level sponsorship for the project, someone who’s got enough clout to keep it going. You want to avoid the kick-the-tires scenario. We’re typically working with line-of-business sponsors who are focused on something that’s pretty transformative to the business process or business they’re running.

CIO: There’s a lot of emphasis on your team working with other parts of IBM. Help me to understand for our readers: How does that bear fruit? What would customers expect to see in the way of new products, hardware, software, from other parts of IBM?

Rhodin: A couple things. One of the ways we’re expanding our APIs out to the masses is through IBM’s Bluemix platform, which is our PaaS. In the first two weeks that the APIs were available they were called 1.2 million times, so we have a number of developers starting to build applications in the wild off of those. We have 100 ISVs that have signed into our developer program that are actually in development sandboxes building commercial applications on the platform right now.

A great example of that is Terry Jones. He was the founder of Travelocity, Kayak. He’s built a new company called WayBlazer on the platform in the travel industry space. He’s a good example of someone who came up with a very different idea that required this kind of technology to do something he couldn’t do when he built Kayak and Travelocity.

He’s building the travel planning concierge that helps you think through what you want to do, where you want to go and what you want to do when you get there, as opposed to having to spend your time searching through many Internet sites and curating the information yourself. It curates it for you and makes recommendations and gives you the underlying supporting evidence to back up the advice it’s given you. It’s a really neat application. I think it’s a good exemplar of what a consumer-oriented approach might be and how people will be able to see this kind of technology in the future.

CIO: But when you look across other product lines at IBM, how does it become embodied in other hardware or software?

Rhodin: In the case of WayBlazer it’s a powered-by-IBM Watson app, so we’re doing ingredient branding on the partners that are working on the platform. We’re going to do the same thing with IBM applications. We’ve announced integration of our Watson Discovery Advisor capability into i2’s intelligent law enforcement applications. Think of it as help for detectives to connect the dots between lots of information.

CIO: It’s like ‘Watson Inside’ on the product?

Rhodin: Yes. And we’re doing the same thing with Kenexa, which is our HR solutions application. You start to see this coming out in multiple applications across IBM.

CIO: If customers want to try this in the cloud, if they want to take advantage of that sandbox capability, what do they have to do?

Rhodin: If you’re an ISV building an app, you have to sign a development agreement with us and you get access to your own sandbox. There is a filter on that because I’m actually dedicating resources and compute time to those partners along with training, etc. But if you’re just a developer that wants to get started on it, jump on Bluemix right now.

CIO: I want to talk about the group itself. Why is it so important that you do things differently than what’s been called Big IBM, whether it’s the job titles or the way HR works or the location? Why is all that important from a customer perspective?

Rhodin: Culture, right. I’m coming at it from a speed and culture viewpoint. I’m very focused on an organization that is looking to the future, thinking differently, that doesn’t feel tied to the way things have always been done.

I’m intentionally cutting strings so that [people] have the freedom of action to try new ideas, to do things differently. A simple little example, the business card in front of you looks different than every other IBM business card. Even the eight-bar logo on the back is slightly off-kilter. We did that intentionally right before we launched the group in January.

Partly, I did it just to prove to my team that we could. The one thing you could never change was the IBM business card, except we did. I used it in a symbolic nature to say that it’s okay to test the boundaries. It’s okay to be bold. It’s okay to do things differently. You have permission. That was the logic on what we’ve been doing.

Ginni [IBM CEO Ginni Rometty] felt that it was important that this new enterprise had its own representative headquarters, something new and different and iconic and very intentionally in the middle of Silicon Alley. Not in midtown, not down on Wall Street, but in a part of town where all the startups are, because those are the people I need to work with.

CIO: What is the one big thing you’ve got to focus on to make this work?

Rhodin: It’s actually very simple. There’s a formula that I was taught a long time ago by Steve Mills [IBM SVP, Software and Systems]: Customers, revenue, volume, profit -- CRVP. I’m at the stage where it’s customers. Are we doing things with enough customers in enough different ways to test and mature the technology and advance it at the accelerated pace we need to run the business?

We’re in customer acquisition mode. We started the year with a handful, that brave handful of customers that came out of the market validation phase and we’ve been doubling our install base every quarter. It’s very rapid growth.

Then the thing that keeps me awake is: How well are we executing on what we agreed to do for those clients, because that will build the references that lead to the next several hundred.

CIO: Finally, what’s the one thing you really want the traditional IBM customer base to know about Watson?

Rhodin: That we’re open and ready for business. Watson is mature enough for real-world solutions. Our biggest Discovery Advisor system right now has over 60 million documents in it. We’re open and can scale. It’s ready to go. Give us a call.

CIO: Speaking of open, would you ever open-source any of the Watson technology?

Rhodin: I don’t know if we’ll get into open-sourcing core technology. The algorithms are moving way too fast and they are the secret sauce of how it works. We have been fairly open with the techniques that were used in the Jeopardy match. There have been research papers written and studied in universities, etc.

We haven’t been hiding things but there is some magic in the way it was put together and that is the basis for competitive advantage. The key thing on that last point is we are opening up the APIs and making them freely available for people within those open-source environments to leverage and use.

This story, "How IBM’s Watson could change your business" was originally published by CIO.

Join the Network World communities on Facebook and LinkedIn to comment on topics that are top of mind.
Related:

Copyright © 2014 IDG Communications, Inc.

1 2 Page 2
Page 2 of 2
IT Salary Survey: The results are in