BHASHINI Sangam workshop in Nepal advances India-Nepal cooperation on multilingual AI
The Kathmandu workshop brought government, academia, researchers and technology stakeholders together to explore language datasets, voice-first AI and locally relevant applications in Nepali, Maithili and other languages.
During the workshop, the BHASHINI team demonstrated multilingual AI capabilities covering speech recognition, real-time translation, voice-enabled interaction and document processing.
KATHMANDU: The Digital India BHASHINI Division (DIBD), in collaboration with the Embassy of India in Kathmandu and Kathmandu University, has organised the BHASHINI Sangam: International Engagement Workshop in Nepal to advance cooperation between India and Nepal in multilingual and voice-first artificial intelligence.
The workshop brought together stakeholders from government, academia, research, technology, innovation and language communities to explore how Language AI can improve access to digital services, information and opportunities in people’s preferred languages.
The engagement builds on a Memorandum of Understanding between DIBD and Kathmandu University for collaboration in language technology and digital public infrastructure. Discussions focused on taking the partnership forward through the co-creation of technology, datasets and locally relevant applications addressing the linguistic needs of communities in both countries.
The initiative is part of BHASHINI Sangam, an international engagement programme of DIBD that seeks to bring governments, academics, researchers, developers, startups, language experts and communities together to develop Language AI across linguistic and geographical boundaries.
The programme places particular emphasis on combining technological capabilities with local linguistic and cultural knowledge. It also seeks to support the preservation of languages and cultures by ensuring that linguistic communities are represented in emerging AI technologies.
Demonstrating multilingual AI capabilities
During the workshop, the BHASHINI team demonstrated multilingual AI capabilities covering speech recognition, real-time translation, voice-enabled interaction and document processing.
The demonstrations highlighted how Language AI can make digital interaction more natural while helping people access digital services in languages they are comfortable using.
A dedicated BhashaDaan session focused on the importance of language data and community participation in building representative AI. Participants explored opportunities to contribute speech, text and image data.
The workshop also examined potential applications in Nepali and other languages, bringing together BHASHINI’s technology ecosystem and Kathmandu University’s academic and research capabilities.
Focus on local languages
Suman Shekhar, First Secretary at the Embassy of India in Kathmandu, said the MoU between DIBD and Kathmandu University provides a strong foundation for taking the collaboration forward.
He said the two sides have an opportunity to develop local applications in Nepali, Maithili and other languages that can create value for people across India and Nepal.
Amitabh Nag, CEO of the Digital India BHASHINI Division, said India’s and Nepal’s shared languages, communities and cultural connections create an opportunity to build Language AI collaboratively.
The objective, he said, extends beyond making technology understand languages such as Nepali, Newari and Maithili to building technology with these languages, their communities and their speakers.
The workshop brought together researchers, students, language experts, developers, startups and ecosystem partners to exchange ideas around language datasets, research and real-world Language AI applications.
The BHASHINI Sangam engagement in Nepal thus seeks to move India-Nepal cooperation in multilingual AI from technology sharing towards a collaborative ecosystem involving technology, academia, government and language communities.
BHASHINI’s scale
The Digital India BHASHINI Division operates under the Digital India Corporation and the Ministry of Electronics and Information Technology (MeitY). According to the release, the platform supports 36 Indian text languages, 23 Indian voice languages and 35 international languages.
BHASHINI currently powers more than 800 government websites, has enabled more than 9 billion cumulative AI inferences and processes more than 24 million AI inferences every day, according to the release.
The initiative positions multilingual AI as an important component of digital inclusion, with the India-Nepal collaboration providing an opportunity to extend that approach across borders and linguistic communities.
