Lightning Talk - Ruuh: A Conversational Social Agent

Abstract

Dialogue systems and conversational agents are becoming increasingly popular in the modern society but building an agent capable of holding intelligent conversation with its users is a challenging problem for artificial intelligence. In this talk, we share challenges and learnings from our journey of building a deep learning based conversational social agent called ”Ruuh” (m.me/Ruuh) developed by a team at Microsoft India to converse on a wide range of topics. The authors are co-creators of Ruuh and the original paper was presented in NeurIPS 2018 Demonstration Track by two of the authors. As a social agent, Ruuh needs to think beyond the utilitarian notion of merely generating ”relevant” responses and meet a wider range of user social needs. The agent also needs to detect and respond to abusive language, sensitive topics and trolling behavior of users. Some of the above objectives pose significant research challenges in the areas of NLP, IR and AI. Our agent has interacted with over 2 million real world users till date which has generated over 150 million user conversations. We intend to walk the audience through our journey of overcoming several research challenges to become the most popular social agent in India.