Like Alexa & Siri, Government plans AI voice bot in regional languageshttps://indianexpress.com/article/technology/tech-news-technology/like-alexa-siri-government-plans-ai-voice-bot-in-regional-languages-5361590/

Like Alexa & Siri, Government plans AI voice bot in regional languages

The move by the National e-Governance Division (NeGD) of the IT ministry is aimed at enabling voice-enabled services that leverage artificial intelligence to widen the coverage for government services offered through its bouquet UMANG app.

Like Alexa & Siri, Government plans AI voice bot in regional languages
How do we get India’s productivity to spike in 10 years? (Illustration: Shyam)

The Centre is planning to develop a voice-based bot powered by artificial intelligence on the lines of Apple’s Siri and Amazon’s Alexa to facilitate access to government services such as utility bill payments, provident fund balance and income tax filings for regional language speakers and the illiterate.

The move by the National e-Governance Division (NeGD) of the IT ministry is aimed at enabling voice-enabled services that leverage artificial intelligence to widen the coverage for government services for smartphone users offered through its bouquet app — the Unified Mobile App for New-Age Governance (UMANG).

To explore the available solutions, the NeGD is expected to “brainstorm” with technology vendors on October 4. The UMANG app currently offers 300 services from 66 central departments and 17 states, including services such as bill payments for water, electricity, gas, mobile, broadband, apart from crop insurance, soil health card services, agricultural advisories and weather forecast services.

As of September 8, UMANG had 73 lakh downloads on Android, 4.31 lakh on iOS and around 41,000 times on the Windows platform.

Advertising

“Currently, the (UMANG) app targets smartphone users assuming these users are able to read and therefore, are able to avail the services offered through the app. However, recent developments have shown that in addition to icons and text-based interactions with users, advanced apps are providing voice-based interactions. This is largely to enhance user experience, and therefore the usage reach of the app to a much larger base,” says the brief for the October 4 workshop.

“In this light, UMANG app envisages to enable support for voice-based interactions with the end-user, especially for people less educated, blind or using vernacular languages, and who may face difficulty in entering text in UMANG application and are more comfortable in using voice interaction.”

According to the brief, the voice bot facility is expected to help the UMANG app move beyond urban centres. “The next wave of growth of internet users is likely to come from tier 2 cities and the majority of these current internet users are native/vernacular language speakers which is further increasing. Given the complexities in typing on vernacular keyboards, voice will become a preferred interface for new users,” the brief noted.

While the solutions are expected to be explored in detail at the workshop, the NeGD has laid down some basic requirements for the voice service. The brief points out that the facility should be such that AI-powered voice bots can be developed and trained to specialise in interactions between the user and the government along with integrating tools such as speech-to-text.

Within the UMANG app, the AI platform should be able to convert user utterances to actionable data based on intents. The brief uses Google’s Dialogflow platform – a framework for interaction technologies based on natural language conversations – as an illustration, to explain the concept of user interactions with the UMANG app using voice.

Private companies are already pushing for voice-based services to push growth in smaller parts of the country. Microsoft is leveraging AI to enable a Hindi language user to communicate with computers and work with digital tools across its Windows 10 and Office 365 products.

Online retailer Flipkart last month acquired an artificial intelligence-based speech recognition start-up Liv.ai, which has developed a speech-to-text platform that supports 10 Indian languages. Flipkart expects the acquisition to help the company in targeting the next 100-200 million users beyond the large metro cities.