Although my interviews with
software firms were long over, I could not resist bright young entrepreneurs. I
found the Pullurs on a visit to Hyderabad. This column was published in Business Standard of 14 January 2003.
TWO WAIFS IN THE JUNGLE OF ENTERPRISE
Jay Pullur still has to pinch himself. His father was in
the Food Research Institute all his life – a smooth, eventless, riskless
lifelong career. That was the Brahmin ideal – a steady job, a single wife,
sedate progression through the four stages of life. In the mediaeval ages Jay
would have learnt the scriptures in a Muth, become an acolyte of a priest, and
have progressed to become the head of a Muth himself. Such a progression from
earth to heaven is no longer possible today; the next best is to make one’s way
studiously through school and college, take a permanent job and retire amidst a
brood of devout, well behaved grandchildren. Risk-taking is for the flighty
banias. And today, Jay finds himself taking a gigantic risk – though he likes
to tell himself and everyone else that he is taking no risk.
Jay is a 1986 graduate of IIT
Kanpur. Immediately after graduation he joined Wipro, where he worked for ten
years. But he was not fulfilled. He wanted to make products, which was
difficult in Wipro: Wipro was a mason, not a property developer. It was
understandable from Wipro’s point of view: it had a vast market for software services
which brought steady, certain revenue; it had no reason to invest in making
products that might or might not sell. But Jay was an engineer. He saw that
profits in services were bound to be limited: the customer knew the seller’s
cost structure, and could work out how long a job would take. So at best he
would pay a certain margin on costs. Services could never be a gold mine.
Whereas a product, if it became popular, could earn many times its cost.
So he left Wipro in 1997, and sat
down with his brother Vijay to plan a different business. The greatest
resistance came from the wider family: no one in the family had ventured out or
done business before. Business, everyone thought, was none of the family’s
business. Raising money, making promises, going and selling things seemed so
out of character. So Jay worked out this bit of sophistry, that he was not
taking any risk at all. An engineer can see where a technology is heading, and
has only to follow the curve. Even if that leads to failure, it is an experiment
– it teaches something: failure is success. With this bit of obscure
philosophizing, Jay managed to overcome the naysayers, and incorporated Pramati
Technologies in April 1998, raising capital from Citibank, Intel Capital and k1
Ventures amongst others. (Pramati means outstanding intelligence or
understanding.) It entered the market with a web application server – a server
that relied on the internet for connecting computers.
The product Pramati chose was
application servers. Business firms generally employ a central computer called
a central processing unit (CPU) which is shared by a number of workstations;
bigger businesses may use a number of CPUs. The software for making them all
work together is called an application server. Sun is the pioneer in making
hardware for businesses and hence in application servers. It wrote them in
Java, which has become the standard for other servers. There is a Java
Community Process (JCP) organization; it holds a fair, Java I, every year.
There is also a Standards Performance Evaluation Corporation which tests
products against standards. The standard for servers is upgraded from time to
time; the current one is called Java 2 Enterprise Edition (J2EE), but previous
versions continue to be used in older or less powerful software. Sun has
created a Compatibility Test Suite (CTS) to test new software.
Today, Sun packages its
application server with its Solaris 9 computer. But other vendors of servers
alone have come up. BEA is the leader. IBM and Oracle, which make very powerful
computers, have servers to match their hardware. HP and JBoss are other players. The leaders –
BEA, IBM and Oracle – charge $10000-12000 per CPU, and have more expensive
powerful versions. The new competitors – HP, JBoss – give their servers free,
but charge $9000-10000 per CPU for maintenance and support. For a business with
multiple CPUs, server software can become quite expensive. The big competitors
do not want to bring down prices; but BEA and IBM also offer cheaper servlet
engines on the side to the buyers of their application servers. A number of
smaller players, including Apache, Covalent and Red Hat – the vanguard of the
Linux brigade – offer less powerful servers for smaller businesses. Like other
software markets, this market was initially served by branded software from big
sellers which embodied elements to make a switch to alternatives difficult. The
newer players have offered products with lower switching costs; when they fall
to zero, the product may be said to have been commoditized.
This is the market Pramati
decided to enter. Early last year, it began to market Pramati Server 3.0 priced
at $5000 per pair of CPUs, and multiple packs with prices going down to $1200.
Together with the server goes another product, Studio, an Integrated Development
Environment (IDE) which enables the client to migrate from earlier standards
and servers to J2EE 1.3, the latest. It passed 15000 CTS tests. By the end of
the year it had acquired over 50 customers, including Mercari Technologies,
which sells software that enables giant retailers to manage their demand
chains.
The product has already covered
its cost. That cannot be the end, for competitors will react with more
competitive prices and products, and their moves will have to be countered. But
in its 120-person team, Pramati has created a resource worth many times the
investment. Jay Pullur considers this team to be his greatest achievement, and
attributes its quality to the choice of engineers – he says attitude and
aspiration are more important than experience and knowledge – and to the flat,
empowering organization of Pramati. These are the plus points that he says
competitors could not match even if they located themselves in India. He says
that Indian offshoots of multinationals are inevitably controlled from abroad
and cannot achieve the same motivation and productivity. He thinks Indians have
tremendous talent, but fail because of neglect of quality and tolerance of
shoddiness. In Pramati he inculcated pride in the people, and they responded marvelously.
Indian software engineers are now known worldwide, but they are looking to
identify themselves with world-class products. When Pramati first began to go
to Java I, many Indians would come up and say, “Well done, guys! Go ahead! Good
luck!” This thirst for achievement is something that, if harnessed by those who
know what achievement means in various fields, could lead India to the top of
the world.
