One of the first languages most of us started communicating in was our native language. As we grew older, English became a dominant language of communication and the native language would seldom cross the comfort our homes and families. As a result, majority of Indians are in the very least bilingual, and in some cases multilingual. However, the Internet has not grown in this hybrid and versatile manner. Introduced in the 90s, the Internet moonshot through the aughts and is now a way of life. And yet, caters to predominantly English speaking parts of the population. In the past decade, some concentrated efforts have been made to mobilise Indian regional languages on the Internet, following its rapid spread and adoption in areas beyond urban pockets. But a synchronised and deeply collaborative effort is needed to accelerate the next billion users towards the Internet that is truly ready for them, not the other way round. 

This was the premise of the National Language Translation Mission, announced in the 2021-22 Union Budget. INDIAai's Sindhuja Balaji caught up with Abhishek Singh, CEO, MyGov and President & CEO, NeGD, and Arvind Pani, Cofounder and CEO, Reverie Language Technologies, on how regional language development is critical to build an India-forward Internet.

How will the proliferation of regional language technologies enhance the reach of education in India? 

AS: Largely, the content available on the Internet is available in English. In the past few years, one can find content in Hindi and some other regional languages. One of our goals at NeGD is to provide access to information, but more so, in a language that everyone can understand. The National Education Policy emphasises on learning in the native languages. If we don't have digital content available in regional languages, we are creating a different kind of haves and have-nots. Children studying in English medium schools will have access to more knowledge and information as opposed to those studying in regional language medium schools. Moreover, learning in one's native language in formative years is known to be beneficial. The National Education Policy focuses on education and pedagogy being spread through digital technologies in regional languages, and for this, the Internet has to rise up and bridge the existing gaps. This needs investment, technology and a collective effort by all stakeholders. This is a high priority initiative by the government.

Can you elaborate on how regional language inclusivity took centrestage during the pandemic?

AP: The month of March last year is still fresh in my mind, when COVID19 came to India. The pandemic and consequent nationwide lockdown necessitated the need for information to reach really reach every nook and corner of the country. With digital being the primary medium for this level of communication, we realised this cannot be done in English alone. One of the most important sources of information that we came to rely on during those months was the Prime Minister's address. This was relayed in Hindi. While 40% of India's population converses in Hindi, the rest speak an array of languages. So we converted these videos, with voiceovers in multiple languages. We have executed two major language initiatives using our proprietary technologies - we have helped develop the COVID section of the MyGov website in 10 languages, and we created a telemedicine helpline for the various states in India in three languages - English, Hindi and the native language of that particular state. A process that would ideally take 3 days, took us about 3 hours. We used our voice tech to modify the helpline scripts and make them available for IVR usage. This ensured people were getting the information fast and in a language they preferred. 

Voice technology is now gaining momentum. How will this impact the development of regional languages and how can this movement be taken to the next line of digital adopters in India? 

AS: In India, we have nearly 500mn Internet users. To ensure the next 500 million can use the Internet with ease, we have to go beyond the traditional models. Not only do we have to build an Internet for the masses in a language they follow, but also empower them with voice technology. Voice is a basic and simple form of communication, and doesn't depend on literacy - which is another barrier to using the Internet in its current form. What we envision is building the next Alexa or Siri but for accessing public services. In the next 3-6 months, we are trying to bring together a host of government and state services on the super app UMANG, with voice assisted technology. UMANG already has has over 20,000 services but we realised that accessing the various elements in the app was a challenge. We are now working on a voice interface that can seamlessly bring together users and queries. For instance, if someone wants to procure blood, one can do so by accessing Rakht Kosh. What we will now make possible is getting this information in seconds by entering a voice command, instead of manually searching the vast database for a specific blood type. UMANG is available in 12 to 13 languages, and nearly 50 services wil be available via voice on UMANG including weather information, passport services, RTO services and more. We foresee a different and exponential growth in the number of people utilising e-services and we encourage startups will offer value-added services above these layers of existing data. Now users can book railway and train tickets, and eventually, can even procure a COVID vaccination too. There is an unmet demand by citizens for services, and accessibility was the missing link. With voice, we can address this lacuna. This is our goal, and its not easy so its success will hinge on partnerships and collaborations. 

AP: I just want to add to what Mr Singh just spoke about. India is a demographically unique and varied nation. Apart from language diversity, we are talking making the Internet inclusive. Language is one way to achieve this goal. We have to consider that a large part of India's population isn't literate and converse in regional languages for their daily needs. voice is a complimentary tech to make customer experience richer and more inclusive. By using voice technologies to augment an online shopping experience, we are catering to a wider set of users and providing them an incentive to benefit from digital technologies. 

What are the challenges one faces while raising adoption levels and enhancing the calibre of research in regional language technology development? How can these gaps be addressed? 

AP: There is a gap that exists in the way we think aboug the design of the Internet itself. With the number of regional language users rising online, it is time to step back a bit and see what a design for India really means. Developers and engineers should be thinking of ways to design digital properties from the ground up for regional language users. What's happening is digital products, services and solutions are being built for English speakers, and are modelled or fashioned to suit the needs of non-English speakers. This doesn't translate to a smooth user experience because sometimes the context is lost in translation. And it ends up deepening the digital divide. With entities like MyGov working on making govt services more accessible, the crucible delivering the digital experience too has to change at a fundamental level. For instance, nearly a billion Indians have Aadhaar cards, so would it be possible to link one's Aadhaar to a language preference? This can be linked to APIs so whenever a service is accessed, the user is welcomed in his language of preference and the entire transaction takes place in this manner. Existing social barriers that have spilled into the digital domain have to be addressed mindfully. 

How is the National Language Translation Mission poised to empower the next billion Internet users in India? 

AS: NLTM has been planned with a vision to achieve almost all the points that Arvind has talked about and a wishlist to create a digital breakthrough to enable communication for regional languages through voice on the Internet. We're hoping to extend voice applications in e-commerce to sectors like healthcare. The government wants to offer a platform with tools and datasets to the general public. Three key initiatives, as part of the NLTM, called Padho India, Likho India and Suno India - aim to harness datasets for three major functions entailing education supported by voice technologies. There are an array of challenges to be addressed, which include getting datasets for diverse dialects, younger languages, and even tribal languages. This will be a crowdsourcing exercise to collect datasets for analysis, following which they will be made public. Many other tools developed by the govt will also be made public for startups and companies to build solutions on top of. NLTM will be a common artefact like UPI and Aadhar, for Indian languages. The government is working to ensure there's harmony across the ecosystem to produce a robust language infrastructure fit for the digital age. 

Listen to the entire conversation on INDIAai's Youtube Channel here.