When 49-year-old Pradeep Sindhu quit his job at rarefied tech ivory-tower Xerox PARC and founded Juniper Networks, most people didn't give the company, or his future as an entrepreneur any chance. After all, Sindhu, an alumnus of Indian Institute of Technology, Kanpur, and Carnegie Mellon's School of Computer Sciences, was a technologist, not a businessman, and Juniper was going head-to-head with networking Goliath Cisco. Still, Juniper's tech-promise attracted fellow IIT-K alum and, arguably, the US venture capitalist with the best record, Vinod Khosla. Today, Juniper is a profitable (albeit, marginally so, according to its results for the first quarter of 2003) entity with a significant presence in the communications service provider market, and Sindhu, its CTO. Did he feel a pang when giving up his post as CEO after seven months of the company's existence? Not at all, claims Sindhu who was in India recently. He knew he didn't have what it took. Excerpts from an interview with Business Today's R. Sukumar.

Juniper was always considered the one company that could give Cisco a run for its money. Has it lived up to that promise?

The only thing I care to live up to is not whether I give Cisco a run for its money but whether we satisfy customers-service providers. We have been, historically, and stay today, the only company that has focused singularly on the service provider segment. All of our competitors who play in that space have other fish to fry, and bigger fish to fry. Our success comes from that focus. We understand the problems service providers have, we build technologies to specifically solve those problems, and use the feedback from the deployment of that technology to build things for the future.

The past couple of years haven't been very good for service providers. The Indian telecommunications market has grown, but telcos in the US and Europe have not done too well. Has that affected your business? Are telcos open to investing in upgrading or expanding their networks today? To us it looks like all they are interested in now is consolidation

My view is that the right way to look at the service provider market is to take a longer-term view. And if you do that, there is no question that at the end of the day there will be a single consolidated technology used to carry all communication. There is also no question that this technology will be based on ip. I think what you saw happen during the bubble and the effect of it is normal given the fact that the technology we are dealing with is not that old. And when things grow very rapidly they are difficult to manage.

That business model had, in some cases, too much optimism. (It involved) in some cases, spending money on infrastructure that wasn't able to give back returns in a timely way-for example 3g and 3g licences. Lots of companies spent money on that without seeing the commensurate return. This is more in Europe. In the US, what people did was, they overspent on the bottom layers of the infrastructure used for communication-the optical layer.

"The right way to look at the service provider market is to take a longer-term view"

The fact is, if you want to have the service provider serve their customer you need to build the network in a balanced manner. This includes not just optical gear but the infrastructure on top of that so that you can add value and you can solve your customer's problems in a cost effective way. All those chickens are now coming home to roost and I suspect that much of that consolidation you mentioned is well into its way to happening.

The interesting thing to note is that throughout this difficult period, the rate of growth of bandwidth hasn't slowed down-even in the worst of times it was 80 per cent a year. That, to me, says that there is huge demand value for the technology that Juniper supplies. The issue that is before the telecom industry is that earnings are not growing at the same rate. That's the thing that needs to be fixed. Juniper has a very strong view on how that can be done. The equipment that we have and the new model called mint, a model for integrated network transformation, will allow carriers to spend money on equipment in such a way that they can actually become profitable and grow their earnings.

It is not anymore about just raw bandwidth and moving boxes, although our competitors would love it to be so. It is very important to us that our customers are successful, and the only way they can be successful is if they are making money and are profitable.

When do you see 3G rollouts starting to happen?

That's a hard question to answer. There are technologies other than 3g, like Wi-fi that are taking the sweet spot. The kind of mobility that Wi-fi offers is very suited to computing. With 3g there's still some distance between the promises that were made and the delivery. Combine that with the fact that there is no compelling application for data.... Once these applications come about I have no doubt that 3g will be useful. The trouble that I have is that I don't see the applications that drive the market.

Isn't that a problem with networks in general? Lots of bandwidth and few killer-apps...

People have poured lot of money into the lower layers of the network, the optical layer and the transmission layer. Those layers are necessary but not sufficient. The value of a network really comes from what I call any-to-any connectivity. No matter where I am, I am able to access the information I want to access as if I were right next to it.

It's a fact of communication that the value of a network increases rapidly with the degree of connectivity. By laying optical fibre here, there, and everywhere one doesn't get that kind of any-to-any connectivity. That is provided by the layer just on top, the data layer. That is provided by IP.

Historically-this is before Juniper arrived on the scene-IP was still there but it was roughly chewing gum and baling wire, which means it kind of worked. You could not simulate the kind of applications enterprises needed for rock-solid connectivity.

"It's difficult right now, but technology will come back into fashion"

Juniper has changed the industry by developing equipment which is custom designed to solve the problem. You end up with networks that are more reliable. You end up with networks that are more secure. And you end up with networks that actually make money.

So, while I agree with you that in the US people spent a lot of money on the bottom layer, the notion that there is plenty of bandwidth floating around of the useful kind, of the any-to-any kind is not exactly true. That kind of bandwidth is pretty scarce.

How do you look at India? One, as a market, and two, as a resource base.

I have a special relationship with India. For me it is very difficult to separate emotion from hard fact. Fortunately, I think this is one case where the two are aligned. India is a potentially large market for telecom-based solely on the numbers. It is also a very fast growing market. And with deregulation happening, that potential for growth is actually starting to be realised.

Within the company, there is actually a significant percentage of people who are of Indian origin. We look at the talent pool that exists in India as a potential for development, at some point in the future. The talent pool is very large and impressive and many global companies have exploited that to their benefit. We have just started to be successful in India. We opened our office in 2000 and have signed on 15-17 customers.

You founded Juniper in 1996. How easy would it be doing that now?

I think it is very difficult for a new company today to enter the service provider market. In a sense, the timing for Juniper was very good. It came about at a time when there was incredible hunger. Networks were literally melting. Things have changed. The bar that a new company would have to overcome is a lot higher. The equipment has improved from what it was. Plus, I think, the carriers are not in a mode where they are willing to experiment with lots and lots of things.

The optimism around technology has got dampened a lot. That has made it much more difficult to start high technology companies today. It's part of the cycle. My own view on technology is to take the long-term (view). As long as you are solving interesting business problems there will be value for that technology in the long term. It's difficult right now, but technology will always come back into fashion. For us (smiles) it never goes out of fashion.

This thing you mentioned about using technology to solve business problems. Was that what was missing at your earlier employer, Xerox PARC?

I had a lot of fun at Xerox. It was a wonderful place- lots of very bright people. And lots of innovation happened there. What was actually missing was the conduit from those ideas to industrialising them. That step, in my own view, can only be taken by people who invent the technology. You can't delegate that industrialisation of an idea to some other person. Industrialising an idea is as important as inventing it. Many people learn that lesson at their own expense. If it is a really good idea, someone else will capitalise on it. Perfect example: Ethernet, Windows-based computing, laser printing, Adobe. There are so many ideas that came out of PARC that Xerox, unfortunately for it, was not able to capitalise on.

Do you think, then, that it is important for inventors to have a strong business focus?

I think it is exceedingly important for inventors to understand that business focus is necessary. It is not so important for the individual to have a business background. Because if you are smart, and you understand that you lack those capabilities, you will ally yourself with people who have that capability.

That's what you have done, isn't it? You do not have a business background...

I do not have a business background and people had given me a zero chance of success because of that. But I clearly understood that the idea needed to be successful from a business standpoint. And where I did not have the capacities myself, I made sure that I had people next to me who did.