India is a country separated by language, and this has been a cause of numerous challenges. To overcome the language barrier, The Ministry of Electronics and IT introduced Digital India Bhashini, powered with the help of AI. 

Bhashini, the National Language Technology Mission (NLTM), was launched by the Prime Minister in July 2022 to provide language technology solutions as digital public goods through the BHASHINI platform. This is aimed to be achieved by harnessing emerging technologies such as AI/ML and NLP to develop and share open-source models, tools and solutions (products and services) for Indian languages with ecosystem development involving startups, industries, academia, research groups, enthusiasts and State/Central Governments.  

The approach is to create large datasets in the Scheduled Indian Languages to train AI models for translation and conversion of speech to text and vice versa, including voice-to-voice translation for general usage and specific domains/contexts like education, healthcare and financial services etc. 

1000+ pre-trained AI models have been made available on the BHASHINI platform. These AI language models have also been exposed via Open BHASHINI APIs for Bhashini ecosystem partners. 

The Bhashini Challenge 

An Independent Business Division (IBD), Digital India Bhashini Division (DIBD), has been set up under Digital India Corporation (DIC) to anchor the Mission BHASHINI activities and to nurture the language technology ecosystem, especially involving startups. In light of the initiative, MeitY has introduced #BHASHINIGrandInnovationChallenge. The participants of the challenge will have to find solutions for two problem statements in the NLP domain: 

Problem 1: Live Speech needs to be simultaneously translated into multiple target languages 

The live speech delivered by the dignitary should be simultaneously translated into local Indian languages for a better understanding of citizens listening to the speech. It should be done in real-time without much delay while the live speech is on. 

Problem 2: Translation into regional languages 

Government of India offices receive multiple communications on paper in regional languages. These documents (both printed and handwritten) are to be digitized using OCR and then translated, then worked upon, and then need to be translated back and responded to in the original regional language. 

The challenge will have three stages. In the ideation and prototype stage, the teams will have to propose innovative and cutting-edge ideas for their solution along with a prototype in 1 Indian language. The top 10 teams would be selected from this stage. Each team will receive funding of INR 1 Lakhs to further enhance the prototype based on BHASHINI APIs. 

This will be followed by an enhancement of the prototype stage, where the shortlisted entries from Stage-1 will get a chance to present their enhanced prototypes to a distinguished Jury in 2 Indian languages. 

Finally, in the solution-building stage, The winner will get a fixed amount of INR 50 Lakhs along with certificate from the Hon’ble Minister of Electronics & IT towards deploying the solution in 10 Indian languages for use by the Government of India and State Governments for a year and further support of Rs 10 lakhs per year towards Operations & Maintenance. 

Things to remember 

The last day for registration is Thursday, 22nd June 2023. The ideas will be analyzed based on the parameters such as approach towards problem-solving, business use case, the technical feasibility of the solution, product roadmap, team ability and culture, and addressable market. 

Find out more about the program here.