'Every application is moving towards a Net architecture'

Thomas Kurian

For somebody who left India for the US when he was 17, along with his twin brother Jacob, armed with only $400 between them, Thomas Kurian, 36, is a completely self made man.

While attending classes, he worked to pay for his Princeton tuition and living costs. He graduated with an electrical engineering degree and a master's in business administration from Stanford University.

Eight years ago, he joined the world's second largest software company straight from consulting firm McKinsey, where he served clients in the software, telecommunications and financial services space.

When Kurian joined Oracle Corporation, its application server software business was a complete mess. Now it is the company's flagship venture, thanks to the boyish-looking Kurian, who is the senior vice president development, Oracle application server. He spoke to Business Standard on the turnaround and the parallels between application servers and databases.

Oracle was one of the late entrants in application servers. But today it is your flagship business. How successful have you been in moving applications from desktop to servers?

Very successful. Every internet website is essentially an application. If you look at how many websites there are in the world, I think they are essentially all applications built on a server architecture as opposed to writing on on our desktop. So we think that this is a fundamental change, it is ongoing and people have really accepted it and shifted to this architecture.

Is the shift evident in non-IT companies?

Absolutely. If you go to most businesses today and ask them what technology they use to share information in the organisation, most of them would say that they use the web and are beginning to use internet technology within the company.

Even the business users of the company who are typically driving the IT organisation do this. And if you think about how many websites people access on a day-to-day basis versus how many applications they start on their desktop computers, outside of your Microsoft Office and e-mail applications, there is very little that you actually do on your desktop.

Most of it is running through the web. Around the world and certainly in India, people have generally accepted that as an undeniable technology trend.

Is it the same for Oracle?

For our application servers, we have 22,000 customers around the world. The majority "� 85 per cent plus "� would be non-IT for us. The application server has been Oracle's fastest growing business in the last three years straight.

The reason why we are seeing this growth is both because of the product offerings which are compelling for customers, from both the feature function as well as the cost of ownership perspective.

Most businesses would want to get into e-business and move to the internet, and the application server is fundamental to enabling that. So most of our customers are companies in the manufacturing industry, chemical industry, pharmaceutical industry, logistics and transportation industries, defence departments, government agencies and education institutions.

So the internet and the e-business transformation has very little to do with IT companies. That's why we are seeing such a large market growth.

How were you able to make this the flagship business for Oracle? What were the challenges?

It's taken three important things to make this possible. On the product side, we recognised a couple of technology trends "� that people were tired of using a lot of middleware to build things.

By giving them an integrated suite of technology, it allowed them to focus on building the application and we took care of all the infrastructure. We recognised that technology trend two years ahead of our competitors, which allowed us to build a suite of compelling technology.

We also hit the market at a time when lots of organisations had recognised the internet, and the internet architecture had lots of benefits and potential for them, independent of the dot com bubble. Every organisation by using internet technology could actually do business more efficiently.

We really had a will to win in this market. We saw a fundamental business opportunity for Oracle, and that led to enormous commitment in terms of sales, resources, support and services, which are all necessary to get customers to use this technology successfully.

Why do companies need application servers? What is it displacing or augmenting?

It's a fundamental question: Does a company want a website, does it want to use it to share information within the organisation? In our view, every application is moving towards an internet architecture.

Everything you do as a company today, you go to work in the morning and you want information on what do you need to do during the day. You then create a business meeting and share something with somebody. You order something, and enter the order on the system. You shift something to somebody and you want to track where it is.

Everyone of these operations that every person in every company does very quickly is moving into the internet architecture. Every one of those systems will be built using an application server. That's why there's such enormous spending in this market segment.

The application server business's growth globally was the fastest, from zero to a billion dollars, in terms of the global spend on any software category in the history of software, in a span of four years. That market now is also conglomerating a bunch of other markets that we thought would be separate.

Like portals and integration or how to access legacy systems were separate. Now, all of these are also converging into this bigger application server market, as a result of which the market is consolidating and there's more segments coming into it.

What kinds of new segments are coming in?

In the past when you built a website, if you wanted to access data out of a legacy system, there was a category of software called EAI "� enterprise application integration. There were companies like Tipco that offered solutions in that space.

Increasingly, application server vendors like Oracle were putting that capability directly into the application server itself. So that market segment which was historically separate is now part of this bigger market.

How different are application servers from web services?

Web services are about specific things. Web services are a technology to get interoperability between different applications. The reason interoperability is important is that people have enormous investments in their existing applications and they don't want to throw them out. So this is where web services come in to allow interoperability.

We have a complete suite of web services technologies in our application server, including an application infrastructure for web services, a way to tie different web services together into business processes, something we call orchestration.

It is a way to configure security for web services, facilities to build web services using our development tools, to monitor and manage web services.

What kind of parallels do you see between databases and application servers?

Very simply, two parallels. In terms of the evolution of the market, traditional databases were created 25 years ago when they were initially created as the killer application.

The thing that drove the creation of the database market was that if you take a lot of data from different systems and put it into a database, you could search for things.

If 20 years ago you bought a database, you typically had one database for transaction processing, a very different one for data warehousing. If you bought a database, you went somewhere else for security and you went to different vendors for clustering, management.

Over time, databases added in a lot of these features and became more mature as an information management platform. Customers got the benefit of lower cost and higher reliability because fewer moving parts mean higher reliability.

Application servers are going in the same direction. This is the technology bet that we made in 2001. The killer application for application servers is that when you run an application on an internet model, you need to support a large number of clients accessing it. It is called scaleability.

So first generation application servers were about deploying your application on a server that supported large number of clients. Now application servers have started incorporating other functionalities.

So to connect your legacy system, you don't need to buy another product "� you use an application server. To build something like a portal, you don't need to build another system but use an application server. You have security across your applications as you are using the same infrastructure.

So in terms of market evolution, these are moving in the same direction which is parallel to the databases.

Then, when you look at databases, in the early part of their lifecycle, people purchased databases for feature functions, and we became the market leader. They purchase Oracle databases today is because of the performance, reliability, high availability and the fact that you can run it on low cost computers using grid etc.

On the application server side, we are seeing similar purchasing criteria from customers, where they are looking for performance, reliability and whether the vendor can support them in fundamental ways.

What is the potential for grid computing?

It is the next wave of standardisation. When I travel around the world and talk to customers, the topic that gets the single biggest piece of attention on both our databases and application servers is grid.

People really see this issue of expensive systems, lots of excess capacity in all companies. So grid is getting a lot of attention and we have thousands of customers who are live on the grid.

And how different is the Radio Frequency Identification (RFID) technology that Oracle offers?

To use RFID, there are two things that we give you that other people don't. The first is that many other vendors give you a specific point solution for RFID.

The problem with that is if you use one vendor's RFID technology today, and tomorrow if another vendor comes with a better technology, you have to change your entire application with each RFID device vendor solution. The market is still very early and who is going to have the dominant RFID technology is uncertain.

We give a software solution that eliminates the need to have a specific dependency on any vendor's hardware for RFID. You can today start with one and tomorrow switch to another, so we protect your investment. That's one big difference.

Second, in order to use RFID, it has to tie into the rest of your business, it has to tie into your ERP (enterprise resource planning) system, into your supply chain management system and it has to tie into your business applications. That is why we have integrated top to bottom RFID with our database, with our application server and our business applications.

In what way did your McKinsey stint help you in this industry?

At McKinsey, you learn about competitive dynamics, how to think of business problems and how to take decisions quickly and effectively using data.

That is fact-based decision making. We entered this application server market, frankly, quite late in 2001. To succeed in this, we had to take a number of fundamental decisions, because when you are late in the market you need to make a few bets and pursue them. So a lot of the knowledge I got from McKinsey helped in that process.

Your brother is the vice president at Cisco. Do you exchange notes with him?

Sure we do. In managing businesses in the technology industry, there are a lot of parallels across these businesses. I learn a lot from him and what he does in running Cisco's storage business. We share notes certainly.