Bhashini: Bridging the Digital Literacy Gap in Tier 2 and Tier 3 Cities Bhasini's Kisan-e-Mitra Chatbot helps people who are not familiar with technology and Nag believes Chatbot has already started bridging the gap

While addressing the issue of building technology that transcends language barriers, Amitabh Nag, CEO of Bhashini, at the Global FinTech Fest, said that Bhashini's (National Language Translation Mission) primary motive is to transcend these barriers. "We are working on Artificial Intelligence (AI) technology to build models that will translate one language into another," he said.

As of now, Bhashini is tackling twelve languages, and the problems it is addressing include automatic speech recognition, text to text and text to speech.

He further added, "We have a lot of printed materials that are not recognizable as digital content but as pictures. Optical character recognition is another problem we are trying to solve. And we have already solved the text-to-text problem in all 22 languages."

While discussing improvements in speech recognition, Nag explained that they have made progress in twelve languages so far, with the goal of reaching twenty-two languages soon. "When we look at this aspect, voice-first enablement means that we are in a position to not only bridge the digital divide but also the literacy divide. Now, the computer or machine can understand and respond to what you want it to do," he highlighted.

Kisan-e-Mitra Chatbot

Bhasini's Kisan-e-Mitra Chatbot helps people who are not familiar with technology and Nag believes Chatbot has already started bridging the gap. While explaining how this chatbot works, Nag said, "For instance, Prime Minister transfers the tranche of money under PM Kisan Samman Nidhi, if a farmer has not received the money, he can ask, 'mera bukhtan nahi hua' (I didn't receive the money), and the chatbot would be able to respond in the same language, 'aapka bukhtan isliye nahi hua kyunki Aadhaar aapka link nahi hai' (You did not receive the money because your Aadhaar is not linked to the account)."

The Chatbot has been working well for RBI, Nag emphasised, now the farmers can have their land records digitally in the state databases, which are in vernacular languages, while banks, which primarily operate in English and Hindi, can still process these records. Previously, if a farmer wanted a loan, he had to take bundles of prints, get them translated in their regional language, and then submit them to the banks—a tedious process that could take one or two months.

"Now, the system transliterates the records. The system pulls the records from the state account, transliterates them, and provides them to the bank—all digitally. So, theoretically, you can grant the loan in a day, provided all the documents and credentials are in order," Nag shared.