Meesho’s PRISM powers AI-led shopping discovery for 264 million users

BENGALURU: If search was once king, discovery is now wearing the crown. That is the bet being placed by Meesho as it reveals the growing impact of its proprietary artificial intelligence engine, PRISM, which is reshaping how millions of Indians browse, discover and purchase products online.

Built for a platform that served 264 million annual transacting users and processed 717 million placed orders in the fourth quarter of FY26 alone, PRISM has become one of the central intelligence systems powering the Meesho marketplace. According to the company, more than 75 per cent of all orders on the platform now originate from AI-driven personalised feeds powered by the technology.

Unlike traditional e-commerce platforms that rely heavily on keyword searches, PRISM has been designed around a different reality. Many consumers, particularly across Bharat, prefer to browse and discover products rather than search for specific items. Meesho says its AI system interprets intent in real time, helping users find relevant products even before they explicitly express what they are looking for.

At the heart of PRISM is a large-scale ranking and intelligence architecture that continuously analyses behavioural, transactional and contextual signals generated during a shopper’s journey. The platform is powered by more than 100 AI ranking models trained on over 400 trillion input signals. Every day, it executes more than 6 trillion inferences within milliseconds to personalise shopping experiences across the platform.

The scale becomes even more apparent during major sale events, when PRISM processes close to 100 million inferences every second to deliver real-time recommendations and product discovery.

The system also supports more than 10 Indian language experiences, including Hindi, Bengali, Marathi, Tamil, Telugu, Kannada, Malayalam, Gujarati, Punjabi and Odia, making it accessible to a broad and diverse user base.

Underlying the entire infrastructure is BharatMLStack, Meesho’s in-house machine learning framework built to handle large-scale AI workloads while reducing inference costs compared with conventional cloud-based systems.

A key component of PRISM is Trendpulse, an LLM-powered system that identifies emerging shopping trends across regions, cities and consumer communities. By analysing changing demand patterns, Trendpulse helps surface products that align with local preferences and evolving consumer behaviour.

Speaking about the technology’s impact, Meesho chief data scientist and head of AI and demand engineering Debdoot Mukherjee said the next wave of internet users will interact with commerce very differently from previous generations.

“The next hundred million Indians coming online will not search, they will discover. They will not type, they will speak, browse, and expect technology to meet them where they are. This is the most significant shift in consumer internet India has seen, and it demands infrastructure built from the ground up for this reality,” he said.

Mukherjee added that PRISM continuously interprets signals across users, products, sellers and browsing journeys, helping commerce move from keyword-led interactions to real-time intent understanding.

“AI in commerce must go beyond efficiency and become something closer to intuition, a system that learns continuously, anticipates intelligently, and makes discovery feel effortless for every Indian. That is the infrastructure we are building for Bharat’s next era of commerce. And we are only at the beginning,” he said.

The effectiveness of PRISM is amplified by the sheer scale of Meesho’s ecosystem. The platform currently serves 263 million monthly active users, records 17 billion product views every day, and has accumulated 1,592 million ratings alongside 505 million reviews. Together, these interactions form one of India’s largest commerce intelligence networks spanning consumers, sellers, creators, product listings and logistics partners.

As online shopping habits continue to evolve beyond the search box, Meesho is positioning PRISM as the engine powering the next phase of discovery-led commerce. With billions of interactions feeding into the system every day, the company believes the future of shopping will be less about typing what consumers want and more about helping them discover it naturally.