Language Processing Using Oracle Digital Assistant

Sales features that require natural language processing, including Ask Oracle bar filtering,rely on the Oracle Digital Assistant (ODA) platform. Here's a brief overview of concepts that will help you complete the required basic ODA setup and integrate the different sales features. Refer to the ODA documentation to understand the full capabilities of the platform. At present, natural language processing is available for English only.

Here's an architecture overview:

1. Oracle Digital Assistant (ODA) is a platform that's used to create chatbots and to provide natural language processing for other applications and features.
2. Different domains, such as sales, service, and financials, require different sets of language skills. Oracle provides some skills as stand-alone components. It builds others into digital assistants. You can install both in your ODA application instance from the ODA "store".
3. For all Oracle cloud application chatbots, and for filtering activities, you install (the term used is "pull") the FADigitalAssistant digital assistant.
4. FADigitalAssistant includes multiple chatbot skills, including CX Sales, Help Desk and HCM Knowledge. For sales, you're using only the CX Sales skill.
5. For each sales feature that uses natural language processing, you create a separate channel. There are different types of channels for different uses.
   • For the CX Sales skill in FADigital Assistant, different channels handle different applications:
     • To enable natural language filtering of activities, you create an Oracle Web channel.
6. After you create and activate a channel, you copy over identifying information from the channel to the sales application to complete the link. You can learn how to create the channel for each feature and how to create the link by reviewing the documentation for the feature.