In the early 1990s, IBM, one of the world's largest technology companies, launched its research arm in India just to provide back-end support to its global business. Two decades on, IBM Research is at the heart of innovation that the New York-based firm is working on.
At its offices in Bengaluru and New Delhi, IBM Research is working to find real-life applications for some of the world's most sophisticated technologies like blockchain and artificial intelligence. Much of this research is focused on solving local problems even as the firm develops products for global clients.
Among other things, the research team headed by Sriram Raghavan, IBM Research's vice-president for India and Singapore, is using artificial intelligence and blockchain to improve crop yield for agriculture companies and helping others transition to the goods and services tax.
The IIT-Madras graduate, who's also the chief technology officer of IBM India, has been with the firm since 2004 and currently focuses on research around blockchain technology.
Raghavan spoke to Quartz about the research centre's evolution in India and its work here.
How has IBM Research evolved in India over the last two decades?
When IBM Research Labs landed in India over 20 years ago (1998), we focused on our delivery centre which provided services to our clients globally. We would service big accounts where we'd get a lot of troubleshooting tickets. The next phase was associated with the telecom revolution in the early 2000s. That's when IBM Research got expertise in data analytics, machine learning, and predictive modelling. We were working with the massive data that our telecom customers were generating and they wanted us to help them make sense of it.
Now, we have evolved into two large, focused areas: artificial intelligence and blockchain. These are the technologies where IBM does new innovations. Outside of North America, the India lab is the biggest research arm focusing on these technologies for IBM. Among other things, at our India lab, we look at how these technologies can be applied to different industries in India.
Could you cite an example of your application work in India?
One area is agriculture. We had done work globally, with a winery, and had a proven example of how the use of water can be optimised through the internet of things, data analytics, and sensors. About 20 months ago, we started looking at India-specific insights.
That led us to ask a question: The average farm holding in India does not have the same purchasing power as big wineries. Does that mean they cannot use digital technologies? No. We have built a collection of artificial intelligence capabilities by combining satellite data and using it to get farmers the same insight that can be gained by sticking sensors in the soil. Now we can tell the farmer about things like the quality of vegetation, health of crops, and moisture in the soil by looking from the sky. This is powerful because I'm not going to have to tell a farmer that he will have to invest in sensorsÃ¢?Â¦because a satellite gives me the same insight.
We've completed three pilots with Indian companies with use cases like pest and disease prediction, improving yield, and yield prediction on India-specific crops like potato and sugarcane.
IBM Research is working with the Indian government on AI and blockchain-related projects. Could you share some details?
On AI and agriculture, we are actually working with the Niti Aayog. They are interested in these capabilities but from the viewpoint of how to drive (in poorer districts) farmer insight and advisory services.
With blockchain, we work closely with the Reserve Bank of India's banking technology. We actually helped them write the first white paper on the technology (blockchain). And it was IBM's blockchain platform called Hyperledger that was used for one of the initial pilots that RBI did.
What are your hiring plans for India?
We started off (in the late '90s) by hiring returning (from the US) talent. But now we hire locally from institutes like the IITs. The advantage for us is we have very strong academic relationships with institutes here, so we work closely with the faculty there.
It's often said that Indian engineering talent is subpar. Is hiring a big challenge for you then?
I think hiring at scale is a challenge, yes. As a research centre, I'm not a scale hirer. I'm looking at hiring the best and only a few people each year. I think we are getting some excellent talent.
What kind of innovation do you think will come out of IBM Research in India over the next five years?
One theme that is very clear to us is that artificial intelligence and blockchain are going to become mainstream in India in a big way. So we are starting to look at what the intersection of these two technologies might look like. It is very clear that blockchain as a fundamental technology for trusted systems is going to have a role in building secure and fair artificial intelligence systems. As more data begin to land on the blockchain, people will want to do more artificial intelligence on it. So how do you work with data on a private, encrypted platform? That's where most of the innovation will come in from.