Ecommerce platform Meesho is increasingly deploying artificial intelligence (AI) across its operations, with nearly 70 per cent of orders coming through its AI-powered personalised ranking system, Prism. In addition, its AI customer support agent, Chorus, currently handles 60 per cent of user queries, said Pratik Kumar, the company’s head of engineering.
While speaking with Business Standard, Kumar said, “AI as a technology is such a powerful innovation that it definitely has unlocked many problems that were otherwise hard to solve. Close to 70 per cent of our orders come through a personalised ranking model via the Prism system.” He oversees technology and AI-related initiatives at Meesho.
Meesho's Prism system uses personalised ranking to recommend products to users based on signals such as demographics, geography, device usage, browsing behaviour, product clicks and ignored listings. The system is designed to identify user preferences even when shoppers arrive on the platform without a clearly defined purchase intent, Kumar said.
The company also uses an AI-powered system called TrendPulse to anticipate demand and help sellers prepare inventory ahead of emerging trends. Kumar cited Navratri as an example of how demand can vary across regions and days, with shoppers looking for particular colours of clothing depending on local customs.
According to Kumar, the ecommerce platform is using AI across product discovery, demand forecasting, customer support, voice-led shopping, fraud prevention, address prediction, logistics and seller enablement.
Meesho is deploying the technology to make shopping more accessible to customers who may be less comfortable typing search queries or navigating text-heavy interfaces. In this case, its Vaani AI agent enables users to discover products through voice, understand product information, add items to their carts and complete transactions through voice-led interactions.
Also Read
On the seller side, AI agents address queries related to financial journeys, payouts, monetisation and campaign creation, reducing dependence on human intervention and helping sellers resolve issues more quickly.
As AI adoption expands, Meesho is also working to contain the cost of running AI models at scale. Its in-house Bharat-MLStack has helped reduce AI inference and model-related costs by around 60-70 per cent compared with external providers, Kumar said.
Meesho currently has approximately 960,000 transacting sellers on its platform, and the company's annual transacting seller base has grown 87 per cent year on year. Here, AI is being used to simplify seller onboarding, particularly for sellers in Tier-II and Tier-III cities.