The Indian AI landscape is experiencing rapid growth, with several homegrown startups leading the charge in developing cutting-edge technology. These companies are focusing on creating AI solutions tailored to India’s unique linguistic and operational challenges, spanning vital sectors like healthcare, customer service, and language technology. Here is a look at five prominent startups, detailing their core focus, their LLMs, and their key capabilities.
Qure.ai
Qure.ai is building artificial intelligence solutions aimed at making healthcare more accessible and equitable worldwide, especially in diagnostics. Their primary focus is on using deep learning and computer vision to analyse medical images like X-rays and CT scans, facilitating the early and accurate detection of various conditions. While the company’s core strength lies in medical imaging AI rather than a conventional text-based LLM, their systems employ sophisticated proprietary deep learning models trained on vast datasets of medical scans and clinical data. These models are capable of instantly identifying signs of diseases like tuberculosis, lung cancer, and stroke, providing an essential ‘second opinion’ for radiologists. Qure.ai’s technology helps prioritise critical cases and works effectively on both new and older machinery, significantly improving patient quality of life through accelerated treatment planning across over 100 countries.
Krutrim AI
Krutrim AI is the creator of India’s first full-stack AI ecosystem, encompassing foundational models, cloud infrastructure, and even custom silicon. They are focused on building AI that is natively multilingual and culturally nuanced for the Indian and global markets. Their flagship LLM is Krutrim-2, a 12 billion-parameter transformer model specifically trained on a massive, high-quality dataset of English, code, and over 22 Indian languages. Krutrim-2 claims to deliver best-in-class performance on Indic language tasks, showing strong abilities in creative writing, summarisation, and complex translation. Crucially, it supports a large 128K token context window, which is ideal for long-form generation and multi-turn conversations, offering a cost-efficient alternative to models much larger in size while deeply understanding the Indian context.
Sarvam AI
Sarvam AI is dedicated to building India’s Sovereign Large Language Model (LLM), having been selected by the Government of India for this crucial initiative. Their mission is to create a bedrock for Sovereign AI for India, focusing on models that are fluent in Indian languages, designed for voice interactions, and capable of complex reasoning. They are developing a family of models, including Sarvam-Large for advanced generation and reasoning, Sarvam-Small for real-time applications, and Sarvam-Edge for compact on-device tasks. These LLMs will be multi-modal and multi-scale, specifically trained to handle the code-mixing and linguistic complexity found in Indian communication. The resulting models will unlock various applications in fintech, healthcare, and e-commerce, allowing for personalised, intuitive, and secure intelligence without data leaving the country.
Hanooman AI
Hanooman AI is a groundbreaking multilingual and multimodal generative AI platform, created through a unique collaboration involving Reliance Industries and the Indian Institutes of Technology (IITs). The platform is building a collection of LLMs, with parameter counts ranging from 1.5 billion up to a massive 40 billion, all rigorously trained in 22 Indian languages. The focus is on providing an open-source alternative to commercial LLMs, catering to crucial sectors like healthcare, governance, financial services, and education. Hanooman’s models boast multimodal capabilities, enabling seamless generation and comprehension of text, speech, and video content. Its immense size and complexity allow it to provide nuanced, contextually relevant, and personalised responses in native languages, aiming to bridge the digital divide for millions of non-English-speaking users.
Yellow AI
Yellow AI is a leading global enterprise AI platform specialising in conversational AI for service automation. Their core focus is on automating customer and employee interactions to deliver a fully autonomous experience across multiple channels like text, voice, and email. The platform’s proprietary LLM is the Orchestrator LLM, an industry-first generative AI-powered agent model that functions with zero initial training. The Orchestrator LLM’s primary capability is its ability to analyse and retain the full context of a customer conversation, recognising multiple intents and seamlessly switching contexts without losing track of the original goal. This allows for human-like, contextual responses and dynamic interactions, leading to massive improvements in customer satisfaction and significant reductions in operational costs for over a thousand global enterprises.
