Home / Educational Content / Oracle Cloud Applications / Building Your Own Chatbot

Building Your Own Chatbot

Shelby Klingerman, Content Curator |

Chatbots are becoming an increasingly popular part of Artificial Intelligence within the workforce. A chatbot is a computer program that conducts a conversation through auditory or textual methods. A recent presentation taught Quest users about what they need to create a chatbot.

HCM Cloud

HCM Cloud is where all of your data is entered and stored. Instead of having to log into your system and search around for the information manually, you could use a chatbot to access the information quicker and easier. However, it is sealed and there’s no direct access to it at a database level. To get data out, you’ll need an Application Programming Interface (API), which are on controlled release within HCM Cloud. You can submit an SR to request API use. With an API, you can either get data out of the system or update it and put it back in. You can set your security levels for these functions within your organization.

Oracle Mobile Cloud enterprise (OMCe)

OMCe is a different channel that allows you to access data within your main system (HCM Cloud). The API within HCM Cloud requires role security to access information. It’s important to be careful about who you give security access to because once they have access, they can see information about any person in HCM Cloud. The intended idea is to provide the role with access to a developer, who can then create a custom API that will restrict and allow information based on the person requesting it, adding back some security that may have been lost. It will handle all calls to and from the HCM API.

Intelligent Bot Component for OMCe

Intelligent Bots utilize natural language processing. They take spoken or written words and convert them to syntax for the OMCe to process. For example, there are several different ways to ask how much money a person makes. You could ask for their pay, salary, earning, etc. The chatbot can recognize all of the different variations and understand the intent behind them—the amount of money a person makes.

To get the bot working, the client asks a question, which is the intent that the bot will pick up on. The chatbot has a number of states for each intent. The intent will pass the syntax (salary, wages, pay, etc.) to the custom API. The bot will then go into a Dialogue Flow that will handle the conversation.

Chat Client

You can use a variety of chat agents with your bot. Some of the options include Facebook Messenger, Alexa or webpage chat. A common question that comes up when talking about chatbots is whether or not to reveal that they are robots. Some Artificial Intelligences have names, like Alexa. Do you let your users believe that they are talking to a real human? Most people would say no and that it’s best to reveal the chatbot for what it is. Ultimately, it’s up to your organization.

Learn More 

To learn more about how these components work together to make a functional chatbot and see a demonstration of how it works, check out the full presentation below.

 

 

Additional Resources

To get more Oracle Cloud implementation tips, subscribe for FREE to the Quest cloud e-newsletter and get weekly updates on the many educational and networking resources Quest delivers to its user community.

Building Your Own Chatbot