Meesho Launches PRISM to Transform AI-Led Product Discovery in India

Meesho has introduced PRISM (Personalised Ranking & Intent Signal Module), an advanced AI-driven system designed to transform how users discover products on its platform. As India’s digital commerce landscape evolves beyond traditional keyword-based searches, PRISM aims to deliver highly personalised, intuitive, and discovery-led shopping experiences for millions of users. With over 264 million annual transacting users, the system marks a significant step toward the future of AI-powered commerce in India.

Shift from Search to Discovery-Led Commerce

Meesho’s PRISM is built around a fundamental shift in online shopping behaviour. Unlike earlier e-commerce models that relied heavily on keyword searches, a large portion of Indian consumers now prefer browsing and exploring products organically.

Recognising this trend, PRISM enables a discovery-first commerce model, where users are presented with relevant products even before they actively search for them. According to the company, more than 75 percent of orders on the platform are now generated through AI-driven personalised feeds powered by PRISM.

This transition highlights a broader change in consumer expectations, where convenience, relevance, and personalisation are becoming central to the shopping experience.

Massive Scale and Real-Time Intelligence

PRISM is designed to operate at an enormous scale, supporting 264 million annual transacting users and handling 717 million orders placed in the fourth quarter of FY26 alone.

The system processes vast amounts of data in real time, interpreting behavioural, transactional, and contextual signals to deliver accurate recommendations. Its architecture includes over 100 AI ranking models that analyse more than 400 trillion input signals.

On a daily basis, PRISM executes over 6 trillion inferences, enabling it to continuously refine recommendations and adapt to changing user preferences. During peak sale events, the system is capable of processing nearly 100 million inferences per second, ensuring seamless performance even under heavy traffic.

Built on BharatMLStack for Scalable AI Infrastructure

At the core of PRISM lies BharatMLStack, Meesho’s in-house machine learning infrastructure platform. This robust foundation allows the company to build, deploy, and scale complex AI models efficiently.

BharatMLStack ensures that PRISM can handle high volumes of data while maintaining speed and accuracy. It also supports continuous learning, enabling the system to improve over time as it processes more user interactions.

This investment in proprietary technology underscores Meesho’s focus on building long-term capabilities rather than relying solely on third-party solutions.

Multilingual and Inclusive Shopping Experience

One of PRISM’s standout features is its ability to cater to India’s diverse linguistic landscape. The system supports more than ten Indian languages, including Hindi, Bengali, Marathi, Tamil, Telugu, Kannada, Malayalam, Gujarati, Punjabi, and Odia.

This multilingual capability ensures that users from different regions can interact with the platform in their preferred language, making online shopping more accessible and inclusive.

As internet penetration expands across smaller towns and rural areas, such features are expected to play a critical role in driving adoption among new users.

Enhancing Seller Growth and Catalogue Intelligence

Beyond improving user experience, PRISM also benefits sellers by enhancing product visibility and discovery. The system analyses product catalogues in depth, ensuring that relevant items are surfaced to the right audience.

It also provides trend intelligence through tools like Trendpulse, a large language model-powered engine that identifies emerging demand patterns across regions and consumer segments.

By understanding what customers are likely to purchase, sellers can optimise their offerings, pricing, and marketing strategies. This creates a more efficient marketplace where both buyers and sellers benefit.

Understanding Consumer Intent in Real Time

PRISM’s core strength lies in its ability to interpret consumer intent without requiring explicit input. Instead of waiting for users to type search queries, the system analyses browsing behaviour, past purchases, and contextual signals to anticipate their needs.

This approach is particularly relevant for India’s next wave of internet users, who may prefer voice-based interactions, visual browsing, or simple navigation over text-heavy searches.

By aligning with these preferences, PRISM ensures that the shopping experience feels natural and intuitive, reducing friction and improving conversion rates.

Leadership Perspective on the Future of Commerce

Meesho’s leadership emphasises that PRISM represents a major shift in how commerce platforms operate in India. The next generation of users is expected to rely less on search and more on discovery, driven by voice, visuals, and personalised feeds.

The company believes that AI systems must evolve beyond efficiency to become intuitive, learning continuously and anticipating user needs. PRISM is designed to meet this expectation by delivering seamless and effortless discovery experiences.

This vision reflects a broader transformation in the digital economy, where technology adapts to users rather than requiring users to adapt to technology.

Handling Scale During Peak Demand

One of the critical challenges for any e-commerce platform is maintaining performance during high-traffic events such as festive sales. PRISM addresses this by leveraging high-speed inference capabilities and robust infrastructure.

During peak periods, the system processes nearly 100 million inferences per second, ensuring that users continue to receive personalised recommendations without delays.

This capability not only enhances user experience but also maximises sales opportunities during crucial business periods.

Ecosystem Scale and Engagement Metrics

PRISM operates within a vast ecosystem that includes approximately 263 million monthly active users. The platform records around 17 billion product views per day, reflecting high levels of engagement.

In addition, the system processes data from over 1,592 million ratings and 505 million reviews. These inputs play a crucial role in refining recommendations and improving the overall quality of product discovery.

Such scale provides PRISM with a rich dataset, enabling it to deliver highly accurate and relevant suggestions to users.

Future Outlook for AI in Indian E-Commerce

Meesho’s PRISM highlights the growing importance of artificial intelligence in shaping the future of e-commerce in India. As more users come online, platforms will need to adopt advanced technologies to meet diverse needs and preferences.

The shift toward discovery-led commerce is expected to accelerate, driven by increasing smartphone usage, regional language adoption, and evolving consumer behaviour.

With its strong technological foundation and focus on inclusivity, PRISM positions Meesho to lead this transformation and capture the next wave of growth in India’s digital marketplace.