Flipkart is leveraging generative artificial intelligence and creating a multi homepage, where buyers can right swipe into a very clean chat-led home page, said Jeyandran Venugopal, chief product and technology officer of Flipkart.
Venugopal, in an event, Converge @ Walmart, said, they are using Gen AI in a large way in the area of customer experience. “In customer experience area, we are summarising the customer experience chat between the users and agents, so that the context is passed to someone else very quickly.”
He added, “And we are running sentimental analyses on the summarised versions of the chats. If the model decides that a user at the end of the conversation is not very satisfied, then it can proactively notify an L2 person to review it and take proactive measures with the customers.”
Gen AI was the most engaging thing in Flipkart’s board meeting in Singapore, added Venugopal. “The most engaging was the Gen AI session. It is a transformative technology that can be used to democratise e-commerce in India. Lot of tier 2 and tier 3 users are new to e-commerce. They need a lot of assisted buying experience. More than any other part of the world, India will definitely have a noticeable impact from Gen AI.”
“We think of Gen AI in terms of three broad pillars. One is experience facing our consumers and sellers and how do we create our chat led experience for them. The other one is core shopping funnel for them. How do we up-level the lot of the work that we do on search for example using Gen AI tech. Third is internal tools for our associates and employees – to be able to improve their productivity levels, to do more with less.”
“On the first pillar, in fact for big billion days and for festive seasons, we are creating a multi home page where you can right swipe and swipe into a very clean chat led home page. For the very first time, we are completely changing the home page of flipkart to be chat led.”
Flipkart has been working on it for last six months and it has shown some highly promising results for the company. Similarly on the sellers’ side also it is doing similar products.
“On the core shopping funnel, if you look at search, it can significantly transform from the existing information trivial based techniques to moving to semantics search and lot of semantics search is underneath powered by large language models.” Now buyers can use any image in combination with any text to search, added Venugopal.