For a country known for its IT and software talent, India’s absence in databases stands out. While Oracle, Microsoft Azure SQL, MongoDB, and Redis shaped how modern applications store and process data, no comparable database company has emerged from India. This raises deeper questions about funding, risk-taking, and the country’s innovation ecosystem.
In a wide-ranging conversation with AIM, Jayant Haritsa, professor at the Indian Institute of Science (IISc), explained why this gap exists and why closing it is far harder than it appears.
Drawing on history, systems engineering realities, and industry behaviour, Haritsa made a point that databases are not just software products but long-term infrastructure bets that demand deep expertise, patience, and trust.
A Historical Asymmetry in Knowledge
According to Haritsa, the roots of the problem lie in how the global software industry evolved before 2000.
“Until about 2000, all the major database companies were located in the US or Europe,” he explained. “The kind of work that they would send to India was work at the periphery.”
Indian engineers were tasked with testing, documentation, localisation, or building applications around databases and not with building the database engines themselves. As a result, the core intellectual capital behind database internals never fully developed in India during those formative years.
According to him, what was missing was not talent, but exposure to the deepest layers of systems engineering.
Building a Database Is Not Using One
Haritsa used a striking analogy to explain the difference between consuming technology and creating it. “Everybody in India knows how to drive a car,” he said. “But can I build a Ferrari car in India? That’s much, much harder.”
Database engineers don’t just work with queries and schemas. They need a deep understanding of operating systems, hardware architecture, memory management, concurrency, and distributed systems.
“Building database systems requires understanding not just databases, but operating systems and computer architecture, because we deal with both the hardware and software platforms,” Haritsa said.
This kind of systems-level thinking only began to spread in India after 2000, when faculty trained in the US returned, and Indian universities started teaching what was “under the hood.”
Over the past two decades, that deficit has narrowed.
Haritsa mentioned examples where India now plays a central role in global database development. Teams in Pune worked on early query optimisers, later absorbed by SAP. Microsoft’s Azure SQL development, he shared, is now largely done out of Bengaluru.
But trust inside multinational corporations does not automatically translate into Indian product companies.
The Ecosystem Problem
Even when Indian teams attempted to build database systems in the 1990s, most efforts failed not because the core engine didn’t work, but because databases are ecosystems rather than standalone products.
“The tricky part with database systems is that even if you build the engine, users expect a lot of supporting tools,” Haritsa explained.
These include schema designers, index advisors, query optimisers, monitoring tools, backup systems, and performance guarantees. Without them, an engine is unusable in real-world deployments.
“It’s like saying I give you the Ferrari engine—build the car around it,” he said. “You need the steering wheel, the brakes, the wheels—everything.”
That means millions of lines of code, years of testing, and sustained investment. Unlike consumer apps or SaaS tools, database companies cannot be built quickly.
“You need deep pockets,” Haritsa said plainly. “This is not something that can be done over a few years.”
Enterprise databases power air traffic control systems, banking infrastructure, and national payment rails. Failure is not an option.
Risk Aversion Against New Entrants
Databases sit at the heart of mission-critical systems, and CIOs are deeply conservative in choosing them.
“If I say I’m a new database system and something goes wrong tomorrow,” Haritsa noted, “everybody will criticise me and say you made a mess by using untested software.”
Choosing Oracle or Microsoft SQL Server is seen as a safe decision, even if it is expensive, because no one can be blamed for following global norms.
He explained that if an Indian database succeeds, it is applauded, but if it fails, the decision to trust it is quickly questioned. This asymmetry makes customer acquisition extraordinarily difficult for new database vendors.
Selling From India to the World Is Still Hard
Haritsa said that most database buying decisions still happen in the US.
“For an Indian company to be able to sell in the US is very difficult,” he said, adding that India historically produced services giants rather than product companies.
While this is changing slowly, breaking into global infrastructure markets requires ironclad guarantees, long-term support, and credibility built over many years—not startup-style speed.
A Decade-Long Marathon, Not a Sprint
So, when will India have its own database company? “It will take at least a decade to develop a very good database system with all the peripheral tools and robustness guarantees,” Haritsa said.
He stressed that success requires engineers who understand operating systems, architecture, distributed protocols, and end-to-end customer behaviour—combined with leadership from industry, not just academia.
“This is a marathon, not a 100-metre dash,” he added.
Looking Ahead: Quantum and LLMs as Game Changers
Interestingly, Haritsa believes the next decade could offer India a rare reset opportunity.
“There are two big game changers,” he said. “Quantum is a hardware game changer, and LLMs are a software game changer.”
Rather than copying existing systems, India could leapfrog by building databases designed from the ground up for quantum architectures and AI-driven interaction, while still preserving the non-negotiable guarantees of performance and robustness.
In Haritsa’s view, India finally has what it lacked for decades: expertise, capital, academic depth, and industry experience.
“All the ingredients are coming together,” he said. If India succeeds, it will not be because it built a faster database, but because it is committed to building infrastructure that the world can trust, over the long term.
The post Why India Still Doesn’t Have Its Own Database Company appeared first on Analytics India Magazine.
