The correct answer is Azure AI Language, which includes the Question Answering capability (previously known as QnA Maker). According to the Microsoft Azure AI Fundamentals (AI-900) study guide and Microsoft Learn documentation, the Azure AI Language service can be used to create a knowledge base from frequently asked questions (FAQ) and other structured or semi-structured text sources.
This service allows developers to build intelligent applications that can understand and respond to user questions in natural language by referencing prebuilt or custom knowledge bases. The Question Answering feature extracts pairs of questions and answers from documents, websites, or manually entered data and uses them to construct a searchable knowledge base. This knowledge base can then be integrated with Azure Bot Service or other conversational platforms to create interactive, self-service chatbots.
Here’s how it works:
Developers upload FAQ documents, URLs, or structured content.
Azure AI Language processes the content and identifies logical question-answer pairs.
The model stores these pairs in a knowledge base that can be queried by user input.
When users ask questions, the model finds the best matching answer using natural language understanding techniques.
In contrast:
Azure AI Document Intelligence (Form Recognizer) is used to extract structured data from forms and documents, not to create FAQ knowledge bases.
Azure AI Bot Service is for managing and deploying conversational bots but does not generate knowledge bases.
Microsoft Bot Framework SDK provides tools for building conversational logic but still requires a knowledge source like Question Answering from Azure AI Language.
Therefore, the service that can create a knowledge base from FAQ content is Azure AI Language.
