Swiggy, the popular food delivery service, is trying to revolutionise customer interactions through Generative AI, allowing more intuitive ordering experiences.
Not sure what to get for lunch? Generative AI can help
Taking on the challenge of deciding what to eat from countless options, Swiggy is using ‘neural search’ to ease this task for its users. This function enables users to use casual, open-ended queries, offering tailored recommendations without needing specific keywords.
Consider this: you type, “I want a healthy lunch after my workout,” or “suggest vegan starters.” Swiggy’s neural search handles such conversational requests, offering personalised choices. It uses a Large Language Model (LLM), tweaked to comprehend food-related terms and Swiggy-specific search data. It responds in real-time to relevant food-related queries accurately.
Swiggy is testing this feature in September and hopes to extend it to all search traffic on the app, based on results. The feature will support voice-based queries and select Indian languages soon, broadening its reach.
A Heavy Sprinkle of Generative AI Everywhere Else
Generative AI also helps Swiggy enrich its catalogue with images and detailed descriptions of items, making ordering less confusing. It’s also integrated into Swiggy Instamart, allowing customers to find groceries and household items conversationally.
Get Ring The Bell, Benzinga India’s weekly briefing. Designed specifically for investors like you.
The application of Generative AI extends to Swiggy Dineout, where a conversational bot guides users to restaurants meeting their preferences. Further, Swiggy is developing a GPT-4 powered chatbot with a third party for enhanced customer service.
Not only consumers, but Swiggy also aims to better serve its restaurant and delivery partners with AI. An AI-powered conversational assistant is in the works that will enable restaurant owners to self-serve, leading to faster problem-solving.
© 2023 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.