The agricultural AI startup KissanAI has announced a collaboration with the United Nations Development Programme (UNDP) to develop a groundbreaking voice-based AI agricultural CoPilot in local languages to support Climate Resilient Agriculture (CRA) practices.

The partnership aims to provide basic advisory services to thousands of Indian farmers, particularly small-scale farmers disproportionately affected by climate change.

One of the main goals of this initiative is to extend the benefits of the AI revolution to a wider audience, especially focusing on reaching a significant number of female farmers. By leveraging voice-based technology, CoPilot aims to overcome language barriers and provide easily accessible, user-friendly information, enabling female farmers to actively engage in and benefit from the progress of AI.

This is the second product launched by KissanAI, following the recent release of Dhenu 1.0. Dhenu 1.0 is a large-scale language model specifically designed for Indian agricultural practices. It is worth noting that this bilingual model excels in understanding queries in English, Hindi, and mixed English-Hindi languages, directly meeting the language needs of farmers.

Dhenu 1.0 has been meticulously trained using a high-quality dataset tailored for Indian agricultural practices. The uniqueness of this model lies in its bilingual capabilities, allowing it to handle a diverse set of 300,000 instructions while supporting both English and Hindi.

Dhenu has not yet been launched in the market as it is currently in the evaluation phase. "The goal is to conduct some Red-Amber-Green (RAG) testing on the model. And the evaluation may take time as our team is small," shared Pratik Desai, the founder of KissanAI, in an exclusive interview with AIM.