Microsoft Introduces Jugalbandi: AI-Chatbot for Rural India’s Farmers

New Delhi: Microsoft has introduced ‘Jugalbandi,’ an AI-driven chatbot designed for farmers and rural users in India. Developed in collaboration with Microsoft Research, AI4Bharat, and OpenNyAI, Jugalbandi operates on WhatsApp and focuses on providing information about government schemes in multiple languages.

The primary objective of Jugalbandi is to bridge the language barrier that exists in India, where English is spoken by only 11% of the population. By offering information in local languages through a mobile phone, Jugalbandi aims to make government programs more accessible to a wider audience.

The chatbot utilizes advanced natural language processing capabilities to understand users’ queries in their native language and provide reliable information, even if the relevant data is in a different language stored in a database.

To interact with Jugalbandi, users can simply send a “hi” message to a designated WhatsApp number. The chatbot then transcribes the text or audio message using AI4Bharat’s speech recognition model and translates it into English using the Bhashini translation model.

Jugalbandi retrieves information on the relevant government scheme using the Azure OpenAI Service’s model and translates the answer back into Hindi. The synthesized response is then sent to the user via WhatsApp, providing them with the necessary information in a language they understand.

Currently, Jugalbandi covers 10 of India’s 22 official languages and provides information on approximately 171 government programs. Microsoft is actively seeking feedback from organizations such as Gram Vaani to refine the application and resolve any existing issues.

The ultimate goal of Jugalbandi is to enable users to understand and enroll in basic government schemes effortlessly. Microsoft envisions a streamlined process where individuals can access government programs hassle-free, promoting inclusivity and empowering users across rural India.

You might also like

Comments are closed.