Meesho Rewires Digital Commerce With AI

Meesho is integrating artificial intelligence directly into the operational fabric of its e-commerce marketplace to prepare for autonomous digital commerce. Rather than treating generative AI as a standalone feature, the company is embedding the technology across product discovery, seller operations, and customer support systems. According to Debdoot Mukherjee, the company’s Chief Data Scientist and Head of AI, the strategy relies on combining large language models with proprietary marketplace data to solve complex commerce problems.

Building proprietary tools for the market drives the company’s data-first approach. Internal products include PRISM, a personalization engine, and Vaani, a multilingual conversational shopping assistant. Meesho has also developed TrustMesh, an AI-driven fraud detection platform. These tools aim to handle the unique challenges of the Indian market, such as diverse address formats.

Related: Top executive moves of the week

Artificial intelligence has also become a fundamental part of Meesho’s internal engineering stack. Pratik Kumar, the company’s Head of Engineering, stated that AI contributes to more than 70% of the code written across the organization. To support this scale, Meesho built BharatMLStack, an internal platform that simplifies the deployment and monitoring of AI models. This infrastructure allows the company to move AI from isolated pilots into reliable, cost-effective production environments.

While the technology allows for high levels of automation, Mukherjee noted that full autonomy requires orchestration frameworks and human oversight. The company expects AI agents to eventually handle complete workflows, from catalog creation and pricing to marketing campaigns. However, he emphasized that these systems will be designed to assist rather than replace human decision-making. This mirrors a broader trend where retailers use automation to streamline routine tasks while maintaining a human connection at key touchpoints.

Related: Nokia Lumia May Make a Surprise Return

Meesho’s ability to execute this strategy is largely attributed to the scale of its proprietary data. The marketplace currently serves 274 million annual transacting users and works with over 1.04 million sellers. Mukherjee argued that the real competitive advantage is not the specific AI model used, but the continuous learning derived from billions of marketplace interactions. This data continuously sharpens the company’s understanding of how users in India shop, sell, and transact.

The company evaluates every model against both technical performance and measurable business outcomes before deployment. Mukherjee stated that a model only ships if it generates a return, rather than just being technically impressive. This focus on ROI ties every AI initiative directly to business results, such as improvements in customer experience and operational efficiency. As the company looks forward, it believes the future of commerce will be defined by intent rather than interfaces, allowing customers to simply express what they need without worrying about how to handle a search engine.