TRANSFORMING BUSINESSES TO DIGITAL

CHATBOT PLATFORM

This document describes the features of the chatbot platform implemented recently for a leading leisure theme-park business in India. It also highlights the chatbot capabilities. “ANDY” is the name of the chatbot.
 
In the current version this chatbot has been developed as Flow-Based Chatbot.
Development work is in progress for the feature-enhanced NLP-Based Chatbot.

In the current version of Flow-Based chatbot following functionalities have been implemented.

• The chatbot asks the user pre-determined questions and provide multiple choices as response which user selects.
• The flow based chatbot is proactive, i.e. the chatbot initializes the conversation.
• The conversation is guided by the chatbot.
• Flow of the conversation can be changed via Content Management System (CMS)
• The questions and their respective choices can be setup via CMS.

In the NLP-Based version which is being developed, following ML based functionalities have been included.

• NLP based chatbot makes use of ML based training models for conversing users.
• The conversation is two way between the chatbot and users, i.e. Both the users and chatbot are active participants in the conversation.
• It converses seamlessly across multiple digital channels and retains data and context for a seamless experience. In best cases, passes that information to a live agent if needed.
• Designed to integrate with critical systems and orchestrate workflows inside and outside of a third party system.
• It can handle real-time action as routine as a password change, all the way through a complex multi-step workflow spanning multiple applications.
• This chatbot can perform complex reasoning without human intervention.

• Andy is Omnipresent – Can be integrated with websites and mobile apps.
• Flow Driven – Currently Andy interacts with the user through a guided approach
• CMS – The questions and their options asked by Andy can be setup using a centralised CMS.
• Andy’s avatar can be changed too using this CMS.
• Booking Engine Integrated – Andy is integrated with theme-park’s Booking Engine.
• A guest/user can quickly and smoothly book a ticket.
• Lead Generation – Andy can ask a user for his mobile number and use it as a lead in future.
• Lead Conversion – Andy can also trace a user’s pattern of chat flow and maintain a report which can be used in lead conversion.

Enhancements in NLP-Based version

• Implementation of NLP capabilities to understand the context of a conversation and to make conversations between Andy and users more linear, seamless and natural.
• Andy can be pre trained to understand the brand specific knowledge and terms to resolve customer requests.
• Andy will be able to recommend or suggest products (Park tickets) for a user based on previous interactions.
• Andy will be converted into a full-fledged virtual customer service executive who will be able to book tickets, cancel tickets, resolve guests’ grievances and handle guests’ queries.
• Data analytics can be used to understand users’/guests’ behavior and train Andy accordingly using NLP.