5-minute interview Dr. Preet Deep Singh

“Government efficiency is needed to provide services to 1.4 billion people. We need  AI for that. There's no other human way to do it otherwise.” 

Learn how the Indian Government is increasing its bureaucratic efficiency with the help of AI from Dr. Preet Deep Singh, VP of AI at Invest India. In this week's episode, Dr. Preet delves into the various applications and examples where AI is helping both the Indian Government and its people live more efficient and easier lives.   

Tell us a little bit about yourself:

So Invest India is a body under the Ministry of Commerce and Industry for the Government of India. I am the Vice President of Artificial Intelligence Adoption. So our task is to see how the Government can increase AI adoption. We see multiple avenues for that. One is leveraging structured data, so let's say, can you get insights from a table so that a person doesn't have to see it? So for a data scientist and AI/ML scientist, it's very easy to see a table and make sense of it. But you're not into data, you can't really make sense of a large table. This person doesn't know what's good, what's bad, and what doesn't work. So my job is to translate gov. to tech and tech to gov., because these are two different entities that work in their own way. And it sometimes requires somebody else to bridge that gap. So I bridge that gap and represent the buy side of the AI ecosystem. 

Is this a field that you've been working in for a long time?

Actually, I started in the field quite recently. So we set up the AI team sometime around 2022. This is also close to when ChatGPT came in, so it became a craze sometime around that. So I think timing wise, we were very good. We got a CEO who's a hardcore technologist. She was with Intel India for almost 30 years, she led it for eight years before joining Invest India. So because she's got a tech mindset and she understands the government, she had this vision that we need to really enable gov. with tech. So from that, AI has fallen into my lap and I am taking care of it. I've been working in AI for a while, but it’s strange, no matter how long I've been working in AI, if I don't look at Twitter for maybe every day or so, I am going to be just lost. So much just keeps happening and so many new developments keep coming in and then you don't know what's happening. So I think a better way to phrase this question would be to ask if you are regularly working in AI. Because that is what's going to make you relevant. 

What are some of the primary applications of AI in your field of work? 

So India is the most populous nation in the world. That means we've got 1.4 billion people who have some kind of interface with the government. We've got birth certificates, identity documents, scheme benefits etc. Because it's tech, it's possible that somebody couldn't get through to something. Somebody's requiring certain services, somebody requires certain help. Solving this at scale is probably only possible with AI. So for example, there is a district collector who gets about 3000 people walking into their office every day. Now this person can't humanly listen to everybody, can't write down what it is and can't sort it. Can we enable it in a way so that people can do it, let's say over WhatsApp. Everyone in India is not literate, but we use WhatsApp a lot because we can send voice notes. So could a collectors office have a WhatsApp number where you just say “I am Rik, I am standing here and I see that there is a pothole on this road and I think that's a risk”? AI can pick up relevant stuff, it creates that in a form and it files it to the relevant department in the district and the collector gets a dashboard. 

Then an investor wants to come invest, but you know India's really huge. So how does an investor decide where to invest? They'll probably ask what is the infrastructure there? What are the policies? Can there be a scheme? Do I get any benefits from the government?  AI can help you answer these questions for each and every investor. It's actually a simple on-prem RAG model that will give you this data from the public domain to a form that is more consumable for you. The same can be applied to train bookings and railway websites. So the scale that we are all talking about here for all of these examples is only made possible with AI. Some of the applications that I've told you about are already deployed and we're trying to scale up the others. 

Why do you think this is such an important field?

Government efficiency is needed to provide services to 1.4 billion people. We need  AI for that. There's no other human way to do it otherwise. You can't replicate the same mind in every place, but you can have copies of a common communal intelligence that comes in through AI where you feed all this data.

Do you have any interesting resources to recommend?

I really like reading Isaac Asimov, his writing is a wonderful foundation for where we are today. It's so surprising that a lot of geeks of yesterday are the AI founders of today, because these geeks were reading those books. So a lot of things that we keep doing are very related to that. But, my advice to everybody is to use GenAI. Use it for whatever you do and see how it goes and try to get the output you have in mind, that's the only way you read somewhere. Otherwise I am no expert. This wasn't around 2 years ago so no one has 10 years experience in GenAI. You can come up to speed maybe by giving it 5 hours, 10 hours a month and that's nothing. To learn a new tech, that's going to change our lives. 

‍

Listen to the full episode: