In this fast-paced world, customers demand ease and efficiency when they talk to a company. Here comes Chatbot, an automated conversational agent which conducts conversations via text or voice. Its main purpose is to strealine interactions between people and services around-the-clock. Chatbots are beneficial for both parties: developing chatbots is cheaper than training and hiring human customer service agents for the company, and customers often prefer a brisk mobile interaction over talking with someone in person or with the call center. From Apple's Siri to Amazon's Alexa, chatbots are making appearance everywhere. However, there are plenty of 'dumb' chatbots in the market that utilised rather straight-forward pattern matching or rule-based approaches.

For this talk, I will be starting with the current state of **Conversational Intelligence** and some common Python libraries used in building chatbots. Various approaches of building conversation engine such as **pattern matching**, **word embedding** and **long short-term memory (LSTM)** models will be discussed. At the same time, I would present the next generation of **Conversational AI** that focuses on **Question Answering** and perform a live-demo of such system. After the demo, the inner workings will be explained and relevant resources (incl a Python framework for conversational AI research and datasets) will be introduced to the participants. Lastly, I would also share my experience launching commercial chatbots with Fortune 500 clients and a few pitfalls one should be aware of before concluding the talk.