The Hindu Business Line : Doing IT differently

Earlier this year, Tata Consultancy Services launched a Co-Innovation Network, a collaborative model aimed at taking the company to the next level of innovation and growth. And Forrester Research featured a research study that highlighted that this network "signals a major shift in the innovation strategy of global IT service providers".

In a follow-up to these two developments, eWorld caught up with K. Ananth Krishnan, Chief Technology Officer, TCS, to dwell on these initiatives, in the second and concluding part of this interview. Excerpts:

Can you spell out the scope of the Co-Innovation Network that TCS launched recently?

From my perspective, the Co-Innovation network is an interesting proposition; the fact that we bring together the core innovation system and that we make it available to our customers, as equal partners. Now this is different from the traditional view of research, where I would invest time and energy by myself and at some time I would magically come down from the mountain and say, here is the research that I produced.

Instead, the other approach is to say that I probably know few of the questions, but I don't know all the answers. Let's kind of bring together our problems, bring together the people who have ways of thinking about it in a different way, whether it is academicians, start-up companies, venture capitalists, etc, etc. And jointly develop the core innovation business model.

It is intellectual property that has been co-developed for the customer by TCS and three or four partners and it adds value to the customers' end-business. In a way, it allows us to become part of the customer's business in a way that we are not in right now.

Any examples of work that you have co-developed with partners?

There are things that are still being explored, but I'll give you some generic points. Web 2.O, for example, is something that has swept the customer space. We use some of these ourselves, our children use it more, we do it with Orkut, writing blogs and so on. Companies are trying to figure out what's the best way to use Web2.O in their businesses. Now TCS or anyone else can construct a Web2.O offer and say here is what Web2.O is from our perspective.

But the way we are looking at Web2.O is that it's a core innovation opportunity. Every enterprise is relatively new to this as they have to evolve their own approach in using this either internally or externally. So our Web2.O Lab, which is right here in Chennai, has a framework by which we bring it in as core innovation. We partner with customers or typical research teams or marketing teams. We bring in our Web 2.O technologies and a few of our Web 2.O partners from some universities and so on. We evolve a Web2.O strategy for a customer and it is being customised right now to be extended whatever yields best results to the customer.

Similarly, a few are using Web 2.O internally for replacing their whole intranet knowledge management. The younger workforce is aware of the chat mode, blog mode or bulletin mode and it gives that youthfulness to the workforce. It need not necessarily translate into business benefit. But extending Web2.O to business transactions, lets say its in-house, an insurance company talking to a customer or its distributor through Web 2.O technologies; could be mobile or handset-based, which, again, exposes them to a newer way of doing business.

How do you prioritise your investment in innovative R&D efforts within TCS?

It's done on a 60:30:10 kind of basis, 60 per cent of time, energy, budgets, and people go into what are called derivative innovation. These are the current businesses that we support with tooling or with productivity improvement. So derivative innovation projects would be things around software development, testing, engineering sciences, management, consulting and so on; 30 per cent of investments go onto new platform innovations, these are extensions of the current businesses that we are offering; that is where utility computing comes in, the new business models around BPO will come in and the next phase of SOA (service-oriented architecture) components factories and all that will come in.

There is a certain set of investments that labs do and these are all typically within the next 18 months or so, all these projects will get into the core. Now 10 per cent of the investments are the blue sky, these are not related to today's businesses, these are not driven by today's customer's requirements.

There is the famous quotation from Henry Ford that says, "If I had listened to customers on how to design the Model T, I would have designed a better horse drawn cart." So that 10 per cent is in that space. I deliberately don't listen to customers, I kind of collaborate with customers, and my co-innovation network and find the next big opportunity. So that is where completely new businesses will be built.

There are some examples of blue sky that are already yielding results for TCS, for example bio informatics. That was blue sky for us three years ago. Now we have reached a stage where we are piloting it to certain customers, and we are starting to win completely non-linear deals in that space, where we would deal with a company of doing discovery and supplying our tool sets.

We will do bio informatics, we are not going to do biology or research ourselves but we will enable the companies doing business in that. For example, we are working on malarial parasite research in India to find the next set of drugs that will be effective against plasmodium, which is currently mutated enough and it is resistant to many of the current malaria treatments.

Now it is a serious problem in India, that is what makes malaria an epidemic in India, but with some of these tools, if we can find the drug that is effective against malaria and its mutations, then we can eradicate malaria in the country. This is 10 years away, but we are playing a small role in this process. It's still a radical innovation point, in no means part of our core innovation right now, but it's a play for the future.

Similarly, there is this recent thing we have done, one of our Wireless Reach projects that won the Qualcomm grand prize for the best innovation in mobile technologies. It's a worldwide prize and we won the first prize and this is about taking wireless into agriculture in India. So we have done a pilot in Maharashtra with testing out how it will be of use to a farmer in India. Which proposition can be built around wireless technologies? So it could be about soil, it could be about rainfall, or planting cycles or pricing of crops that will be grown, whole range of things, a few of which can be enabled through wireless technology.

For example, soil and moisture content, which would dictate the way in which planting should be done. It's actually a big question in most areas and then recommend as to which particular crop is best suited for which particular soil, if you have to maximise the outcome and so on. So this pilot has won the grand prize.

It's an opportunity to see whether we can scale it up beyond the one or two districts that we worked it on and see whether it can do something dramatically different. Now if that happens, that potential is another 200 to 250 million users or potential users of mobile technology in the country. So these are all horizon three projects and potentially disruptive projects which could help us. So we make sure that we spend 10 per cent of our time on these kinds of things, we spend 30 per cent of our time on more visible or fixed time line projects and then 60 per cent goes in making sure shot things. That's how we prioritise it.

In terms of testing, you have partnered with Georgia Tech. Several companies have been examining testing as a key focus area with enormous automation possibilities. Extending this to the ADM (application development and maintenance) part of your business, how do you think non-linearity will kick in and what will be the quantum of change?

Two observations from my side. One is to look at what has actually been done. I'm not going to venture into what will be the results in the future, but if I look at the last 12 years, TCS in 1995 was approximately Rs 600 to 700 crore of revenue and in '95 we probably had revenue per person of about 10 to 12 thousand dollars as a rough gross measure of productivity. Today, we are close to $4 billion and our revenue per person would be north of $50 thousand per head, which means that from a past performance perspective, we have increased our productivity by 4x. Now as a CTO, my challenge is to make sure that this is not just past performance. If we look at others (domestic players), they are also doing around $ 50 - 55 thousand per person.

Take a company like Accenture, their revenues are $15-16 billion. They have grown substantially in India, they are going about $130-150 thousand per person as revenues. This means there is a potential of 3x gap in what we do and what the western side (or multinational majors) will do.

Let's strip aside all complexities, their business is different than ours, but this is the end result we have to shoot. Now the way I look at it is that, going back to testing, it's an area that certainly is attractive to a lot of people in the industry today because there is a need to do independent testing, verification and validation, to do unit testing, system testing, all kinds of testing. This is a tactical opportunity and TCS also plays in that and our assurance services are quite good.

But I'm looking at it from the CTO perspective; I have to look at if from the point of what is it that we are trying to achieve in the whole testing exercise. We are trying to assure not just the customer but the customers' customer, that this software is trustworthy, it is doing what it is supposed to do, from a function perspective, performance and security perspective and various other parameters. The second thing we are assuring the customer is, if I make a change in the software, then the overall piece would still function as advertised.

If I take an analogy from a day-to-day function, if you go and get your puncture fixed, you don't necessarily test the radiator for leaks, the rest of the car functions smoothly. While in a car it is easy to visualise that, a tyre is distinct from an engine, but in software we don't know. We would typically say, we fix something in your software module A, but if you really want to be sure, you have to test the car again, and that's what adds to the cost and that's what adds to the complexity of the whole software.

So what we are doing with Georgia Tech, for example, is to try and reach that minimal set of testing which is required to assure the quality of the overall product, without adding to the costs. So in that sense it's disruptive and goes against the grain of what today's state of the art is, which is to retest everything.

I have to test everything and therefore it costs time and money to the end customer, I might be successful in terms of time and money, which I have earned today but in the end it's not a good value proposition back to the customer. If I can go back now to the customer and say I have a way in which I can distinguish between the tyre and the engine, and assure you that if I change the tyre, I just need to test a couple of things around the tyre, not touch anything else in the car, that's attractive.

That's what the team in Georgia Tech working with TCS are trying to evolve, a robust way to assert minimality of testing, so that's one aspect of it. We have also got a few start-up alliances. There's a company called Ajuba with whom we have a core innovation network alliance, which has got a way of assuring quality of unit testing in Java code. Given the volume of Java code that is written, asserting quality of Java code, even as you write the code is a useful proposition, which will result in some advantages later.

These are sort of strategic investments being made by TCS to try and change the way we look at it in the next one or two years and bring in the non-linearity into testing. If I can go back with technology that can assert minimality of quality of testing upfront, I don't have to do as much testing as I need to and I can then pass on that value to the end customer. And obviously there would be a business value to that kind of proposition. That's the way I look at the whole assurance part.

More Stories on : Interview | Software | Tata Consultancy Services Ltd

Article E-Mail :: Comment :: Syndication :: Printer Friendly Page

Stories in this Section
Receive M-gyan

`Open Season' for Java

From NetBeans to bean bags

`Create more product development companies'

Disk partitioning

Recording TV programs on PC

Making the mobile phone a tool that teaches

Doing IT differently

Uncertain world

Quiz

Cartoon