Indian e-commerce firms adopt AI and automation to manage costs

Indian e-commerce and direct-to-consumer (D2C) startups are increasingly adopting specialized AI and automation to manage operational risks and costs. To mitigate the high cost of refused cash-on-delivery orders, some companies use automated software to call customers at late hours to confirm addresses and encourage prepaid payments.

Customer support in the region is heavily centered on WhatsApp, where chatbots serve as the primary layer for handling repetitive tasks such as tracking orders and processing refunds. Successful implementations focus on narrow, well-defined workflows rather than broad, open-ended language models to avoid errors in policy or refund promises.

On the infrastructure side, the e-commerce platform Meesho is prioritizing the development of its own in-house machine learning platform, known as BharatMLStack. By building this foundational stack rather than relying solely on third-party cloud services, Meesho aims to reduce AI workload costs by 60–70%. This strategy is designed to maintain the thin margins necessary for its low-price business model while standardizing intelligence across advertising, fraud prevention, and cataloging.