AI Chatbots Introduction
There are numerous frameworks and propriety off-the-shelf chatbot offerings, but most do not clearly map out the much-needed control of data by an organization. Often the propriety offerings take an organization’s private data for training on the cloud and provide the outcome as a model. In this course, we will focus on data privacy and control over the development process. The chatbot that you will learn how to develop could be completely built in-house using open-source JAVA frameworks and NLP libraries in Python.
The beginning of the course helps you understand the processes in the banking industry and delves into identifying the sources of data to mine for the intent from customer queries. The second part focuses on natural language processing, understanding, and generation, which are demonstrated using Python. These three concepts are the core components of a chatbot. In the final part, you take up the development of a chatbot called IRIS using open-source frameworks written in JAVA.
The following are the key topics this course offers:
- Identify the business processes where chatbots could be used in an industry and suitably guide the design in a solution architecture.
- Focus on building a chatbot for one industry and one use-case, rather than building a ubiquitous and generic chatbot.
- Natural language understanding, processing, and generation
- Learn how to deploy a complete in-house-built chatbot using an open-source technology stack like RASA and Botpress (such chatbots avoid sharing any PIIs with any third-party tools)
- Develop a chatbot called IRIS from scratch by customizing an existing open source chatbot framework.
- Use APIs for chatbot integration with internal data sources.
- Deployment and continuous improvement framework through representational learning
We hope you enjoy the journey.
Textbook: Abhishek Singh, Karthik Ramasubramanian, Shrey Shivam, “Building an Enterprise Chatbot: Work with Protected Enterprise Data Using Open-Source Frameworks”, ISBN 978-1-4842-5034-1, Apress, 2019
0 Comments