Teach ChatGPT to Answer Questions: Using Azure AI Search & Azure OpenAI (Semantic Kernel)

A model cannot answer questions about your documents unless the relevant content reaches it. For a collection of PDFs, that starts with storing, indexing, and retrieving the text needed for a question.

This tutorial uses Semantic Kernel to connect Azure AI Search and Azure OpenAI. I work with PDF documents stored in Azure Blob Storage, then use search results to give the model context for its answer.

The article follows that path from documents to retrieval to generation. It is the Semantic Kernel version of the document-based question-answering workflow I explored in this series.

Read the original article:

Teach ChatGPT to Answer Questions: Using Azure AI Search & Azure OpenAI (Semantic Kernel)