The client wanted to include Chat Support and Automated Conversations into their customer service strategy to make customer support available 24/7 with instant responses in multiple languages basis customer demographics. IGT helped deploy a Chatbot solution via an intelligent conversational assistant called 'Zippy', enabling automated responses on the website and Facebook Messenger to customer FAQs in six different languages.
Share this Post
Airline envisions to include Chat Support along with Automated Conversations into their customer service strategy to make customer support available 24/7 with instant responses in multiple languages basis customer demographics. The aim is to have a more natural communication with customers regarding their travel queries over popular and easy to access messaging channels in their preferred language. To elevate their customer support experience via conversational AI and to provide differentiated customer service, Chatbot solution is adopted knowing that it can help overcome the following challenges-
- Cost-effective and easily scalable compared to voice-driven support
- Fluctuating query volume especially during peak season or disruptions
- Limited agent support concerning subject expertise and language
- Inability to calculate Customer satisfaction score(CSAT)
- Volume Shift - 45% of chats can be handled by Chatbot independently
- 24x7 operations for enhanced customer experience
- Reduced contact center cost for multiple language support & training
- Positive impact on CSAT due to reduced wait time and faster processing
- Availability of information on the go and in written form
About the Client
The client is an upcoming low cost international carrier based out of Tokyo expected to start its operations from May 2020. It is planned to operate medium/long-haul international flights.
- IGT Solutions (IGT) has built and deployed a Chatbot solution via an intelligent conversational assistant called 'Zippy' enabling automated responses on website and Facebook Messenger to customer FAQs in six languages- English, Korean, Japanese, Thai, Simplified and Traditional Chinese. The plan is to extend it to other channels like WhatsApp, KakaoTalk, Line and WeChat soon.
- Customers will be asked few details like name, email id etc. at the start of the conversation following which queries written in natural language text related to check-In, baggage, booking etc. are swiftly handled by the conversational assistant.
- The Chatbot solution uses NLP cognitive service for customer query interpretation and intent matching and fetches the appropriate response from FAQ repository (maintained with over 200 FAQs for each language).
- The Chatbot is enabled with a 'Bot to Agent' handoff feature. In case the customer query is not trained as part of FAQ utterances, the solution will seamlessly route the conversation to a live agent with matched skills. 'Bot to Agent' handoff is enabled by integration with 'LivePerson' chat platform wherein the entire chat history is transferred to ensure the continuation of user interaction from the bot routed conversation.
- When the conversation ends, the customer gets an option to print or email the chat conversation for future reference. A feature of the Post-chat Survey is also provided to capture customer feedback on their interaction with the Chatbot.
- IGT worked with client's business team to design the Chatbot's conversation flow along with other elementary features like Start Over, Welcome/ Greeting Message, Language Change, Fall-back Message, Transfer to agent etc.
- Continuous Bot training will be done by retraining of system logged missed utterances flagged by the Bot. The agent will qualify valid queries to become a part of the training set, which will then be used to improve the Bot's capability to understand several intents. This activity can be performed for all languages in scope by skilled Chat Support agents.
- Reporting parameters are also captured to measure Chatbot's confidence & productivity.
Automation Tool Partner
- Google Dialogflow