February 2020
Intermediate
41m
English
conference
Migrating millions of users from voice- and email-based customer support to a chatbot
by Madhu Gopinathan, Sanjay Mohan
Closed Captioning available in German, English, Spanish, French, Japanese, Korean, Portuguese (Portugal, Brazil), Chinese (Simplified), Chinese (Traditional)
Overview
At MakeMyTrip, a large number of customer care agents were dedicated to handling customer support issues over voice and email. Given the popularity of WhatsApp in India, conversational interfaces have become mainstream for the Indian consumer. MakeMyTrip decided to capitalize on this trend and revamped customer support channels using a chatbot named Myra. Myra is adept at handling the languages of the Indian consumer including English, Hindi, and a mix of the two popularly known as Hinglish.
Madhu Gopinathan and Sanjay Mohan explain the high-level architecture and the business impact Myra created. Myra handles user input using a natural language understanding framework that solves problems such as intent classification and slot extraction using the latest advances in deep learning. In the intent classification problem, “Please cancel my ticket” should be mapped to a label "full cancellation.” The intent classification model supports over 120 intents and is sophisticated enough to distinguish between fine-grained intents such as full cancellation, partial cancellation, and flight cancellation by airlines. In the slot extraction problem, from, “Please cancel my ticket to New York,” the value, “New York” should be extracted to fill the destination slot.
For curating the data required for training deep learning models, MakeMyTrip employed multiple methods, such as manual tagging, creating language models from existing email conversations, progressive refinement of the models, and natural language generation. In this case study, Madhu and Sanjay dive into the insights from developing deep learning models to power Myra, revamping of agent tools, and processes to support Myra and the ensuing business impact.
What you'll learn
- Discover insights from deploying a conversational bot and associated agent systems to improve postsale service
- Dive into architecture, strategies for data curation to train deep learning models, and the business impact from deploying a chatbot based on these models
This session is from the 2019 O'Reilly Strata Conference in New York, NY.
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Watch now
Unlock full access
More than 5,000 organizations count on O’Reilly
O’Reilly covers everything we've got, with content to help us build a world-class technology community, upgrade the capabilities and competencies of our teams, and improve overall team performance as well as their engagement.Julian F.
I wanted to learn C and C++, but it didn't click for me until I picked up an O'Reilly book. When I went on the O’Reilly platform, I was astonished to find all the books there, plus live events and sandboxes so you could play around with the technology.Addison B.
I’ve been on the O’Reilly platform for more than eight years. I use a couple of learning platforms, but I'm on O'Reilly more than anybody else. When you're there, you start learning. I'm never disappointed.Amir M.
I'm always learning. So when I got on to O'Reilly, I was like a kid in a candy store. There are playlists. There are answers. There's on-demand training. It's worth its weight in gold, in terms of what it allows me to do.