🚨 We're Hiring! Join us at AI4Bharat 🚨 📍 Role: Research Analyst 🕒 Tenure: 3 months (extendable based on performance) 💻 Work Type: Remote Are you passionate about India’s diverse cultures, languages, and heritage? Do you love to explore, research, and uncover the

🤖 Curious how LLMs handle multilinguality? Check out RomanLens: The Role Of Latent Romanization In Multilinguality In LLMs at #ACL2025 🇦🇹 poster! 📅: July 30, 11:00 AM 📄: arxiv.org/abs/2502.07424 Let’s talk interpretable ways to understand multilingaulity in LLMs!

Excited to present my work at #ACL2025 today! I'll be at my poster on "Towards Building Large Scale Datasets and State-of-the-Art Automatic Speech Translation Systems for 14 Indian Languages". 🗓️ Today, 11:00-12:30 📍 Poster Session 1, Hall 4/5 Paper: aclanthology.org/2025.acl-long.…

Excited to present my work at Ann-Cathrin Lunke today! I'll be at my poster on "Can Vision-Language Models Evaluate Handwritten Math?". ⌚️Today, 11:00-12:30 📍Poster Session 1, Hall 4/5 📄Paper: aclanthology.org/2025.acl-long.… Come say hello! #Multimodality #HandwritingRecognition #ACL2025NLP

In 2019, when the Chief Justice of India asked for court judgments to be translated into Indian languages, it exposed a much deeper gap: the lack of reliable AI models for most languages. Instead of solving only for the courts, our team set out to create a foundation that would

🚀Meet AI4Bharat INTERSPEECH 2025 ! We’re hosting an informal meetup on Thursday, 21 Aug @ 3 PM 🚀 Top Floor, Food Court Foyer (tables near booths).🚀 Why join? Students & postdocs → explore research opportunities at AI4Bharat Researchers & postdocs → learn about faculty

The scale of IndicVoices is powered by people. Nearly 1,900 coordinators, mobilizers, and experts worked alongside 16,000+ participants to capture speech in 22 Indian languages. Their role wasn’t limited to logistics. They built trust, helped participants navigate apps, and

AI4Bharat began with a handful of students at IIT Madras exploring deep learning and building small models, testing computer vision, and translating text. What started as experiments quickly became a focused mission: making AI work for India’s many languages. Today, AI4Bharat’s

Grateful to be named a recipient of the Google PhD Fellowship 2025 under the NLP track! Thanks to Google and my wonderful AI4Bharat family for making this journey so special.

🚨🚨🚨 New paper : The Reasoning Lingua Franca: A Double-Edged Sword for Multilingual AI. Large Reasoning Models (LRMs) often default to English for multilingual tasks. We compare reasoning in English vs the question’s original language and analyze the trade-offs. 🧵1/N

🎙️ Have you used our ASR or TTS models? We’re seeking feedback from professionals, developers, and researchers using AI4Bharat’s ‘Voice AI’ tools. Your input will help improve future releases and support the ‘Voice AI’ ecosystem for Indian languages. 🔹 Why participate? -