IT, BPM firms log rise in requests for AI chatbot pilots

Companies like Tech Mahindra, Genpact and Fractal Analytics see high interest in the usage of generative AI for content generation in areas of customer support. India can create a differentiated offering in this space with its huge workforce, said the executives, speaking to ET during an industry roundtable on the technology.

For ChatGPT to be useful for businesses and enterprises, the applications need to have a stronger contextual background of the business cases that the enterprise deals with, said Nikhil Malhotra, Tech Mahindra’s chief innovation officer.

“Telecom companies are trying to see whether they can upsell and cross sell using generative AI. These are very small use cases at a proof-of-concept level,” said Malhotra.

Sreekanth Menon, vice president and global leader – AI/ML services, Genpact, said the BPM service provider sees demand for automated content generation solutions as well as solutions where generative AI can model human-like interactions with customers.

“Automated content generation is something which we are seeing in the market where customers are requesting to see how generative AI can reduce the time to generate content and offer intelligent personalised solutions for customers,” said Menon.

Discover the stories of your interest

However, the onus is on companies to build their proprietary data pools running generative AI applications for data security purposes.Srikanth Velamakanni, group chief executive at Fractal Analytics, said while companies have a lot of interest in the solution, it is not feasible for everyone to invest in building their own large language computing models required to run a generative AI platform.

“Maybe you can look at open-sourced large language models, bring it inside your database, tweak it and start to generate solutions with it. So, I think the applications are huge and companies are looking for cost-effective ways to execute it,” said Velamakanni.

In fact, India can create a differentiated offering in this space with its huge workforce and technology companies serving this space, said Sangeeta Gupta, Senior VP, Nasscom.

“Given the fact that we understand and have dealt with issues of data security, privacy, while building the BPO and IT services industries, and because Indian companies are working on a lot of digital transformation programmes, we can bring in that expertise as trustworthy partners for the world to explore generative AI applications,” said Gupta.

Industry experts suggest a process of continuous education of employees and consumers, higher levels of security, innovative watermarking solutions and revisiting frameworks for intellectual property classification among others.

Most importantly, due to the overwhelming popularity of generative AI solutions like ChatGPT across the masses and student population, experts suggest a larger need to understand how the solutions work to prevent the spread of misinformation.

“There is a need to educate people that the solutions are generated by models that run on data and statistics,” said Raghav Aggarwal, cofounder, Fluid AI. “This is particularly important in the case of platforms generating content as people need to understand the risks and challenges associated with using such content and the limits of these solutions,” he added.