Build your private Chatbot with Botkit and Snips NLU: Host your Home Assistant on your computer

Build your private Chatbot with Botkit and SnipsBook descriptionChatbots or conversational agents are increasingly present in messaging applications to be closer to customers or help employees with their daily tasks. Most of the time they are built on solutions hosted in the cloud. The goal of this book is to code an open source chatbot solution hosted locally to ensure the privacy of conversational exchanges. Botkit, for the conversation management and Snips NLU for the natural language understanding, combine to allow us to build such a solution. What you will learnYou will learn how to code a chatbot with Node.js and Botkit for the conversation management and Python for building a Natural Language Understanding service based on Snips NLU library.As in the book on Dialogflow, we propose to build the chatbot by use, that is by starting from a minimal conversation to which we add, step by, step new capabilitiescorresponding to patterns of conversational exchanges and implementing new functionalities of Botkit and Snips NLU. Here are the steps followed in the book:To welcome, to leave the conversation and answer by defaultIdentify and remember whom it is talkingRecognize an intention expressed in natural languageTo connect to a business service to build a useful answerTo chain complementary questions to obtain mandatory parameters (slot-filling)To chain several interrelated questions - conversation memoryInterweave conversation threadsDeploy on the Slack messaging platformTransform the chatbot conversation management program into a web serviceTo conclude, the chatbot is transformed into a web service that can come to enrich a micro-services architecture on which we can build any human-machine interface hosting the messaging locally. This architecture ensures that conversations do not circulate in the cloud but remain internal to the company's information system.Who is this book forA basic programming understanding is sufficient because programs have been reduced to the essentials. Each step gives rise to a complete commented program which is given entirely in appendix so that it can be easily retrieved in a code editor and executed locally on your computer.

Author: Hugues Van eylen

Learn more