Making GenAI work for Indian enterprise IT

Making GenAI work for Indian enterprise IT was the next topic. Sandeep Dutta, Chief Practice Officer & Executive Team member, Fractal Analytics, and Golok Kumar Simli, Principal Advisor & CTO, Global Passport Seva Programme, MEA, were in conversation with Vikram Jeet Singh, Partner, KPMG India.

AI has been here since three decades in various shapes and forms. It has taken a sharp uptake in the last few quarters, thanks to GenAI. There are some fears of jobs going away in the future. GenAI is a path-breaking tech, and also has some concerns. How can we leverage GenAI for Indian enterprises?

Golok Kumar Simli, MEA, said GenAI is an emerging technology. It has been put into many baskets. Our perspective is on reaching scale, of 1.4 billion people. We can use GenAI as a tool, and look at achieving customization for the Indian market. We can have cultural capability, as well. There are several barriers to address. We are making this as a success story. 

Today, anyone can apply for a passport from India. We have achieved that because of the technologies. We have data strategy. We must understand the datasets very well. If I am pumping data to power GenAI, we need to see where we are going wrong. GenAI is best tool for data analysis. We need to train that and allow policy makers to take advanced decisions. We need to read citizen sentiments, and give back to them.

Sandeep Dutta, Fractal Analytics, added that we have been in the forefront of analytics, and help enterprises make better decisions. AI is a transformational technology. With GenAI, the power of what AI can do is becoming more clearer. We are in the early days, and there will be greater impact.

We also need to bring in datasets. There is also a human in the equation. They will be using the tool. There are tools about diabetes, and smokers. We need to bring AI, and also some interventions, to help change behavior. We are developing apps that help you take more steps, for another countrywWe need to think holistically to make a difference for India. Adoption is always faster in a smaller country. We can now do things via GenAI. There are issues around data privacy, testing, etc.

Simli said we can tap knowledge for GenAI. We can have massive change management. We have an DPDP law, and we take consent from the prinicipal owner of the data. We need to have consent of the necessary people. We need to also train the principal owner of the data. We also have data privacy and security concerns. We can learn from the rules of engagement. They all need to be aligned. We need to judiciously harness opportunities.

Dutta noted that we have responsible AI. It is for the industry to bring up. Large tech leaders are recognizing and incorporate that. There will be regulation. We have an AI toolkit for India, to ensure your data is not biased. We need to have a maturity assessment. There are indications to check whether people are paying lip service or are accountable. It can also give interface for users to unlock, in an intelligent, natural-level form. As an enterprise, you also need to look at your data, build guardrails, and invest in that. 

Vikram Jeet Singh said we have a term secure-by-design. In government, we need to start with responsible AI. ChatGPT3 has better controls. Before you launch the next model, you need to demonstrate whether you can have control.