by Dorai Thodla (@dorait) on Thursday, 20 July 2017

-4
Vote on this proposal
Status: Submitted
Technical level
Beginner

Abstract

Conversational Interfaces are gathering momentum. It is difficult to design a good CI. There are two types of interfaces - rule based and AI based. Most of the popular Chatbots use rule based interfaces but we see a slow shift towards AI based intelligent interfaces.

In this session we will look at the challenges in building good conversational interfaces, how AI can help and how Python is one of the most desirable languages to implement these systems.

This talk will cover various topics - conversational interfaces, chatbots, natural language understanding, natural langauge generation, existing tools to develop conversational interfaces like Rasa, Google’s api.ai, Amzon’s Lex and others.

Outline

  1. What are chatbots and why they will impact the future of application interfaces?
  2. The architecture of a Conversational App (aka chatbot)
  3. The Dialog, NLU, NLG loop
  4. Challenges in implementing intelligent conversational interfaces
  5. A brief overview of various tools for building your conversational applications

Requirements

internet, projector

Speaker bio

A technopreneur and a seriral entrepreneur with companies in India and US. Currently dabbling in Applied ML and Chatbots.
I have done variations of the talk several times - for TiE, Nasscom, Monage (conference in US), meetup in Berlin. I have been managing a couple of teams building skill bots, infobots and other chatbots for over a year.