During Quest Experience Week (QXW), we explored the world of chatbots, digital assistants, and the future. There are thousands of chatbots in use today, and if you’ve called or texted a customer services group, then you’ve likely encountered one. It’s important to understand where chatbots are in use and what their value is to both the individual and the organization. There are both simple and more complex use cases for the implementation and use of chatbots and digital assistants within a modern ERP application.

Overview of Chatbots and Digital Assistants

There are several important things to remember when talking about chatbots and digital assistants. Chatbots are not as intelligent as digital assistants, and they lose the context of a conversation. However, chatbots are good at answering general questions. Digital assistants are capable of carrying out a wider range of functions to assist in completing business tasks. Digital assistants are also better at remembering context than chatbots.

Oracle Digital Assistant (ODA) has six key components that make it beneficial.

1. ODA is structured with artificial intelligence (AI) and natural language processing (NLP).
2. APIs connect ODA to ERP and SaaS systems.
3. ODA uses question and answer skills for frequently asked questions (FAQs).
4. ODAs are exposed through chat, voice, mobile apps, and websites.
5. Instant Apps provides forms to capture structured information.
6. Insights provide built-in analytics.

Digital assistants provide consistent service, help cut down on operational costs, and forecast. They are available 24/7 and are scalable to handle high volumes. They execute a friendly, consistent script, and record results and confirm solutions. Digital assistants also offload the workload for simple to moderate repetitive tasks and questions from staff.

In 2018, there were roughly 265 billion support requests that cost $1.3 trillion. According to Chatbot Magazine, digital assistants can save up to 30 percent of that. The forecast is that, by 2020, 80 percent of organizations will use digital assistants, and, by 2025, all applications will have AI.

Digital assistants are widely used at big-name organizations like Boston’s Children’s Hospital, The Wall Street Journal, eBay, WebMD, UNICEF, ATAR, Lyft, and Fidelity Investments.

Chatbots and Digital Assistants in PeopleSoft

Within PeopleSoft, both task-oriented and data-driven/predictive digital assistants are available. You are able to access employee self-service and utilize things such as address change, life event change, tax change, open enrollment, class schedule, looking at grades, or financial aid information.

There are quite a few features that PeopleSoft Digital Assistant delivers and even more that are on the roadmap. The Absence Management chatbot, Absence Assistant, was delivered in HCM Image 31 and separate SKUs for ODA. There are several interesting items on the roadmap, including:

• Basic questions for basic company-related information
• Enrollment in benefits
• Mobile HR assistant
• Onboarding
• The ability to talk to employees
• Interacting with employees during training

When designing the digital experience with digital assistants, it’s important to provide a rich user experience, often through the perception that the user is speaking with a real person, provide better 24/7 service availability, build expertise and development efficiency, and gain organizational buy-in and understanding.

The image below depicts a visual of the governance framework for chatbots and digital assistants.

There are three important aspects to take note of when selecting your first digital assistant process:

1. Choose processes that are done often and by many. Examples of these might include asking questions that have known answers or expense and time reporting.
2. Choose processes with high impact such as life events for benefits or a student checking their grades.
3. Processes should be mature, well defined, simple, and small.

Three things to contemplate when looking at design considerations:

1. Persona
2. Brand
3. Image assets

In regard to persona, perhaps you can partner with your marketing team to identify an archetype to model the digital assistant persona. Archetypes are generally associated with a color. Use this to further build out the brand color palette. In relation to image assets, it is suggested to use the Small Form Factor image for a chat window header and digital assistant icon for the chat window.

It’s also important to note what configuration and development of chatbots and digital assistants entail.

Configuration includes:

• Creating a skill with intents, flows, and custom components
• Creating the digital assistant and attaching the skill
• Creating a channel and linking it with the external application (PeopleSoft requires the Java SDK)
• Refining the conversation flow

Development steps include:

• Configuring and deploying the Java SDK in PeopleSoft
• Creating the custom components
• Web Application/Instant App development

Testing and lifecycle management included four key components. Unit testing, system integration, review, and proactive feedback must all take place. Unit testing requires testing each skill with the Skill Tester and testing the digital assistant with the Skill Tester. System and integration testing require testing the digital assistant via the SDK that was deployed on the external application and testing authentication and validating security requirements. It is also important to review insights analytics for skills. You will need to have proactive feedback gathering to improve both the mechanics and tone.

The next steps for getting started with chatbots and digital assistants include:

• Identify the best use case and create a business justification
• Provision Oracle Digital Assistant in the Cloud
• Design, configure, develop, test, and deploy
• Review and optimize
• Repeat