Flipkart said its newer seller-facing systems can infer product details, recommend attributes and improve listing quality with far fewer manual steps
| Photo Credit:
ANDY
Artificial intelligence has quietly become the operating backbone of India’s e-commerce and quick commerce platforms, moving well beyond search and recommendations to power seller on-boarding, cataloguing, demand forecasting and last-mile execution at scale. What was once a surface-layer feature is now embedded deep in daily operations, helping platforms compress timelines, reduce costs and manage hyperlocal complexity in a market defined by speed and choice.
“At quick commerce speeds, AI becomes the decision layer that ties together demand sensing, inventory placement and last-mile execution,” said Bharath Chinamanthur, SVP – Supply Chain Innovation and Seller Experience at Flipkart. The company uses AI-driven forecasting down to the pin-code level to stock fast-moving items closer to demand hotspots, while machine-learning models dynamically recalibrate inventory and routes through the day to respond to micro-spikes triggered by weather, local events or payday cycles.
On the seller side, AI is reshaping the earliest touchpoints. Flipkart said its newer seller-facing systems can infer product details, recommend attributes and improve listing quality with far fewer manual steps. “We’re also using AI to create product descriptions, review summaries and even generate images, reducing the need for time-consuming and costly photo shoots,” Chinamanthur said, adding that this significantly shortens time-to-list and improves catalogue readiness.
This shift is particularly meaningful for first-generation entrepreneurs in smaller towns. By combining mobile-first onboarding with AI-enabled cataloguing, platforms are lowering barriers linked to confidence, cash flow and technical know-how. “For most small sellers, AI adoption is still platform-led and largely invisible,” said Madhav Kasturia, founder and CEO of e-commerce logistics start-up Zippee. “Sellers aren’t logging into AI dashboards. They’re seeing faster on-boarding, fewer catalogue rejections and better demand signals. In India, AI succeeds when it disappears into the workflow.”
Over the past 12–18 months, AI has also helped standardise messy catalogues and reduce human dependency in day-to-day operations. Image recognition, attribute extraction and duplicate detection are now largely automated, with human oversight reserved for edge cases and context-heavy categories. According to Kasturia, the biggest gain has been consistency — lower rejection rates and quicker go-live timelines — allowing sellers to operate across multiple marketplaces with the effort it earlier took to manage just one.
For consumers, AI-led discovery is changing how shopping journeys unfold. Flipkart said it has reimagined search using GenAI-powered semantic and multimodal capabilities that understand intent rather than just keywords, including visual matches. Recommendation systems continuously learn from browsing behaviour, purchase history, region and seasonality, improving product findability and reducing “dead-end” journeys that hurt conversions.
Amazon, which has been scaling AI across its India operations, said the technology now touches every step of the customer journey. “AI improves selection, lowers costs and accelerates delivery — never as an end in itself,” said Rajeev Rastogi, Vice-President–Machine Learning at Amazon. The company uses AI to serve lighter pages on weak connections, enable shopping in eight Indian languages and power its generative AI assistant Rufus, which answers open-ended questions and offers conversational recommendations.
On the operations side, Amazon said AI validates addresses, optimises packaging and improves supply-chain efficiency by about 25 per cent, turning inventory approvals from days to near-instant. “Over 100,000 sellers use our AI tools, and 94 per cent report improved outcomes,” Rastogi said.
As platforms scale, resilience remains critical. When AI systems fail to refresh quickly enough during peak demand, decision latency can ripple through availability signals, substitution logic and routing. To mitigate this, companies are pairing AI-led decisioning with stress testing, dynamic scaling, and human safeguards at critical touchpoints.
Looking ahead, executives say the biggest competitive edge will come from delivering superior customer experience efficiently at scale — combining hyperlocal planning, high-quality discovery across languages and visuals, and trust and safety systems that can catch fraud and low-quality listings early. In a market where India often runs on exceptions, AI’s real test will be how well it adapts to the country’s unique complexity.
Published on December 18, 2025
