Q&A is a product that supports candidates and/or employees by preemptively preparing your AI Assistant for the questions and answers around your company's information, benefits, culture, and more. When a candidate or employee asks your AI Assistant a question, the system will automatically use NLP, or Natural Language Processing, to recognize the person's intent and determine the best topic within the system that matches their question. From there, your AI Assistant will send that matched question's response to the candidate/employee.

NLP is regularly trained by Paradox's Engineering Team to strengthen your assistant's understanding of the conversation, but enhancing Q&A with Contextual Q&A allows advanced language models to serve as your assistant's foundation for conversing and understanding a conversation's context better than ever before.

Elevating your assistant with Contextual Q&A

Contextual Q&A falls under the Contextual AI product and bases its knowledge off of two aspects:

• The knowledge (questions and answers, also known as Q&A topics) added to the Content Management System (CMS)
• Advanced language models that help your assistant understand the context of the conversation and respond more accurately

The combination of the above enhances the conversation quality and ultimately makes the conversation feel more human-like and relevant. Click here to learn more about contextual conversations.

Contextual vs. defined responses

Within the CMS, you can determine whether a Q&A response is Contextual or Defined. From here, your AI Assistant will respond accordingly.

• Contextual responses – The responses that will be adapted to the context of the conversation, using Contextual AI (using LLM). They will likely be changed when sent to a candidate or an employee. This is the default selection when Contextual Conversation is enabled for your account.
• Defined responses – The responses that will be sent exactly how they are saved in the Answer value (using NLP). This is the type of response provided to candidates/employees when the above Product dependencies are not all enabled.

Contextual Q&A + NLP

While rare, if Contextual Q&A does not kick in during a conversation, then NLP will jump in after one minute to ensure that the candidate receives the appropriate response.