Ecom means biz with AI: Technology repurposed to serve Indian customers

As more people in towns and rural areas shop online, firms are localising and training AI tools for cultural nuances

Srikant Sree Ram, director of Amazon Fresh India, stood in a farm in Mahabaleshwar, a hill town in Maharashtra’s Satara district, watching as freshly picked white strawberries made their way to a collection centre nearby. A farmer told visitors that the strawberries are supplied exclusively to Amazon Fresh customers in Pune and Mumbai.

 

As Amazon Fresh, the ecommerce giant’s platform for daily grocery needs, expands to 170 Indian towns and cities, seasonal produce like Alphonso mangoes of Ratnagiri and onions from Nashik are being sourced directly from farmers. What ties the company’s services — procurement to delivery — together is a supply chain powered by artificial intelligence (AI) and machine learning (ML). 

 

AI is reshaping ecommerce: from delivering fresh produce and personalised recommendations to resolving queries in Indian languages. Firms are using AI to improve customer and seller experience, convert queries to sales, and streamline services.

 

Amazon has formed a team of agronomists that gives advice on improving farm yields and quality. ML and computer vision algorithms streamline supply chain processes, and identify defects in fruits and vegetables, reducing waste. It uses ML to give farmers information about their crops, including detecting infestations through its Kisan app. A farm produce undergoes a four-step quality check at collection centres before being sent to processing facilities. 

 

“We've built our operations with a deep appreciation for India's scale and diversity, integrating AI and ML across every touchpoint of our business,” said Rajeev Rastogi, vice-president of machine learning at Amazon. “What excites me most is how these technologies work together to solve distinctly Indian challenges while setting new benchmarks for operational excellence globally.”

 

GenAI for customers

 

Amazon uses generative AI (GenAI) to summarise customer reviews into insights like ‘battery life’ or ‘ease of use’, allowing customers to quickly decide about a product. Customers can tap a highlight to view related reviews, streamlining decision-making.

 

As AI helps ecommerce firms in groceries and farm produces, it is shaping online fashion too. Myntra’s technology helped the Walmart-owned digital fashion retailer to engage with millions of new customers who use features like MyFashionGPT, My Stylist and Maya.

 

MyFashionGPT, powered by ChatGPT, the wildly popular chatbot of OpenAI, helps customers search for clothes and accessories through natural language prompts based on occasions, moods, and celebrity styles. My Stylist, allows users to upload outfit images and receive personalised styling recommendations. Maya enables natural language conversations to simplify shopping and “reduce decision fatigue”, according to Myntra.

 

The company’s Virtual Try-On feature lets users try multiple beauty products, like for eyes, face or lips, simultaneously from their devices. The feature has increased customer conversion rates two-fold, said Myntra.

 

“There are many exciting and unsolved problems across a wide range of functions such as customer support, catalogue, design and search that can be streamlined with AI,” said Lakshminarayan Swaminathan, head of product management and design at Myntra. 

 

The Indian ecommerce market is expected to be worth $325 billion by 2030, growing at a compound annual growth rate of 21 per cent, according to a report by Federation of Indian Chambers of Commerce and Industry and Deloitte. 

 

Diverse needs

 

What makes Indian ecommerce’s AI adoption remarkable is its complexity: models are being trained not just on users’ browsing behaviour and purchase history, but also on the country’s vast linguistic and cultural diversity. As more people in towns and rural areas shop online, companies are localising AI tools to speak the language — literally — of India’s next billion internet users.

 

Ecommerce firm Meesho launched India’s first multilingual GenAI-powered voice bot that offers personalised, human-like assistance in multiple languages. As 80 per cent of Meesho’s users are in Tier-II and smaller cities, the bot is meant for working efficiently on basic smartphones and in noisy places. The bot can distinguish between casual affirmations (like “ji” or “okay”) and genuine interruptions.

 

Separately, the company’s GenAI-driven customer support assists customers in six Indian languages. It resolves about 90 per cent of user queries at one-fifth the cost, achieving higher scores in customer satisfaction compared to human agents, according to Meesho’s 2023-2024 annual report.

 

Quick-commerce firm Zepto uses data science and ML for forecasting demand, acquiring customers, pricing, advertisements, delivery partner allocation, customer support and other work. 

 

“By deeply personalising these systems for individual customers, we've achieved remarkable improvements in both customer satisfaction and business profitability,” said Nikhil Mittal, chief technology officer, Zepto.

 

Ecommerce firm Snapdeal said AI has helped it to improve delivery timeline prediction and courier allocation. "We have been able to resolve over 30 per cent of our customer queries using chatbots,” said a Snapdeal spokesperson. 

 

The next frontier is agentic ecommerce where AI agents will make autonomous purchasing decisions for buyers and sellers, according to a report by investment firm Prosus published in collaboration with data platform Dealroom.co.

 

“AI adoption is no longer optional — it's imperative. Companies leveraging AI are accelerating their relevance to consumers, enhancing personalisation, and driving unprecedented growth,” said Fabricio Bloisi, chief executive officer of Prosus Group, in the report.