Inside Meesho's big AI bet: Why India's most-downloaded shopping app is betting on infrastructure
· Source: expresscomputer.in
In short
Meesho's FY26 annual report, its first as a listed company, outlines an infrastructure-first AI strategy built around its in-house BharatMLStack, which the company says cuts inference and AI workload costs by 60–70% versus equivalent cloud services. Flagship systems include GeoIndia LLM for landmark-based address geocoding (20-point accuracy gain, 5% lower misroute costs), the Network Intelligence System for ML-driven route and truck planning, the Vaani voice shopping agent (1.5 million users in month one, 22% conversion lift), and TrustMesh for fraud and return-to-origin prevention. Meesho reports over 70% of its code is now AI-generated, with 264 million annual transacting users and 2,668 million orders in FY26, while contribution margins compressed as technology and AI infrastructure investment rose.
Read Original Article — Read the full article on the publisher’s site.
In the issue of week 41, 2026: AI Moves Supply Chains From Planning Toward Execution, But Data and Capacity Set the Pace
Tags: logistics, geocoding, personalisation, e-commerce, voice AI, route planning