At its flagship annual event Shiprocket Shivir 2025, India’s leading eCommerce enablement platform, Shiprocket, announced the launch of Shunya.ai, India’s first indigenous large language model (LLM) developed in collaboration with AI startup UltraSafe. Positioned as a truly built-in-India, for-India model, the company says that Shunya is set to revolutionize how Indian businesses and consumers interact with generative AI.
The announcement was made by Saahil Goel, MD and CEO of Shiprocket: “Today I am very, very excited to announce a partnership which is about what India gave to the world – Zero. No computing is possible without zero. No arithmetic is possible without it. It doesn’t mean nothing. Instead, it means potential. I’d like to announce our own Indian LLM called Shunya.”
All About Shunya.ai
Shunya.ai, powered by UltraSafe, is hoping not to be just another generative AI platform. It is tailored for Indian use cases, speaks over 15 Indian languages, and is deeply integrated with the Indian tech ecosystem. Shiprocket says that Shunya is designed for inclusivity and hyper-local relevance including automating tasks for MSMEs, integrating with local software like Tally or marketplaces like Shopify, and supporting Indian language voice commands.
Vishesh Khurana, Advisor – Strategy and Growth, Shiprocket, shared the vision behind the initiative: “When we thought about Shunya, it was about solving Indian problems with Indian tech. Most generative AI tools are copy-paste. They’re not made for our systems or scale. Shunya changes that. Every Indian language is supported, every local platform is integrated. Built by Indians, for India, in India. Just like Shiprocket changed logistics for India, Shunya will change AI for Bharat” From retail logistics to merchant services, Shunya is expected to integrate across Shiprocket’s core operations, thus making AI accessible and practical for Indian sellers.
A Model of Innovation and Sovereignty
Raaid Rahman, CEO and Co-founder, UltraSafe, highlighted the larger implication of the launch: It’s incredible to see the product updates launched today. Indians have significantly contributed to the evolution of AI globally. With Shunya, we finally bring that innovation home. The model and data weights remain in India, ensuring data sovereignty and complete local control.”
Adding to the technical depth, Arpit Sharma, CTO and Co-founder of UltraSafe, explained: “Shunya is India’s first multimodal LLM trained on 18.5 trillion tokens, including 3 trillion tokens in Indian languages. It understands not just text but also images, voice, and video, and is natively agentic, meaning businesses can integrate it with a single line of code without any data leaving India.” He emphasized that Shunya was engineered to respect Indian culture, values, and emotional nuance, bringing unprecedented relevance to enterprise-grade AI in the country.
Shunya.ai App and Voice-Powered Copilot
Saahil Goel announced the release of the Shunya-powered app Shunya.ai, now available on Android, urging users to explore the platform and contribute to its contextual learning: “As more merchants and consumers engage with Shunya.ai, it will keep getting smarter. We’re launching task agents for booking tickets, GST integration, UPI assistance, internet search, and more. Indian needs, Indian problems, Indian solutions.”
For Android users, Shunya.ai is now available for download. With this step, Shiprocket and UltraSafe invite developers, businesses, and citizens to contribute to a growing AI movement—rooted in Indian context, and built for Indian ambition.
Other Important Announcements Made at Shiprocket Shivir
The Shiprocket Shivir also showcased significant AI-first product enhancements. In a live demo, Goel introduced Shiprocket Trends AI, an AI assistant that allows merchants to talk to their data. By asking natural language questions like “Top trending categories in Surat under Rs 1000?” or “High RTO categories in Kolkata?”, users can gain deep insights drawn from billions of transactions processed across the Shiprocket platform. “Dashboards are limiting. But with conversational AI, you can slice and analyze any data the way you want at your fingertips,” Goel said.
Voice-Powered Copilot: Democratizing AI Access
The new voice-enabled Shiprocket Copilot was another key highlight. In a multilingual country where voice is often more intuitive than text, this launch aims to break digital barriers. “Voice is the great democratizer,” Goel explained. “With unstructured workflows and multiple languages, voice AI can bring productivity to the next billion users.”
AI-First Vision for Shiprocket and Bharat
The event also highlighted Shiprocket’s evolution into an AI-first company, with Goel outlining a threefold AI strategy:
- AI for Shiprocket to optimize internal operations;
- AI for Sellers including tools for SMBs and entrepreneurs;
- AI for Bharat democratizing access and building infrastructure for India’s digital future.
“We’ve spent the last 10 years building for you,” Goel told the audience. “Now we’ll build with you, using AI.” The resounding message at Shiprocket Shivir 2025, which was projected by the company was India’s tech ecosystem no longer needs to rely on foreign-built solutions. With Shunya.ai, Shiprocket Trends AI, and Voice Copilot, the company is doubling down on building infrastructure that aligns with Indian languages, workflows, and scale.
