After Infosys, Wipro, and Persistent, Tech Mahindra joins hands with NVIDIA to launch education LLM under Project Indus

Tech Mahindra shares traded 1.5% lower on Friday amid broad-based selling pressure on the software and technology stocks. The shares traded at ₹1,455 apiece on the NSE at 3:00 pm. As NVIDIA signed multiple partnerships with leading IT and technology companies in India, Tech Mahindra was another name that was added in the list.

Tech Mahindra has teamed with NVIDIA to launch a cutting-edge Education Large Language Model (LLM) under Project Indus. Powered by NVIDIA, Project Indus is a Hindi-first LLM built with authentic Indian linguistic and cultural context for sovereign, scalable AI.

The new LLM will debut as an educational model focused on democratising high-quality learning. It is designed to empower millions of students with a deeper foundational understanding of subjects such as physics and other disciplines, while also embracing India’s rich linguistic and cultural diversity. With support from multiple partners, the new LLM aims to be a one-stop platform for educational use cases in India.

The new education LLM has been developed using advanced AI frameworks and infrastructure support from NVIDIA, including the NVIDIA NeMo framework and deployment through NVIDIA NIM microservices, enabling scalability, performance, and production readiness.

To address data scarcity challenges in select languages, the Tech Mahindra team employed innovative techniques, including generating half a billion synthetic tokens with NVIDIA NeMo Data Designer. Additionally, the LLM enables Agentic AI by supporting the creation of autonomous AI agents that can understand and respond fluently in natural Hindi, facilitating AI-driven learning experiences and broader adoption across Hindi-speaking users.

Commenting on the development, Nikhil Malhotra, Chief Innovation Officer and Global Head of AI and Emerging Technologies, said, “ AI is becoming central to national digital infrastructure and inclusive growth, but global foundational models are not designed at country-specific levels. Through the project Indus, our collaboration with NVIDIA directly addresses this gap by delivering Hindi-first, sovereign AI models that are scalable and relevant.