How to link figures in webdocuments and other sources for an LLM chatbot?

Hello,

Currently I have an LLM chatbot running on the sample code from streamlit via this workshop: https://aws.amazon.com/blogs/machine-learning/quickly-build-high-accuracy-generative-ai-applications-on-enterprise-data-using-amazon-kendra-langchain-and-large-language-models/

and using bedrock with RAG approach using this workshop as sample: https://github.com/aws-samples/amazon-bedrock-workshop/blob/main/03_QuestionAnswering/01_qa_w_rag_claude.ipynb

I want the chatbot to respond with relevant figures along with the textual response to the question. To do this. one solution could be to use AWS textract and get the images out and label them with various text groups that would be relevant to that figure. This way when an answer has a text group that is similar enough to the ones associated with the figure, the figure can be selected with a given confidence level as well.

However, I don't see any clear pipelines to extract various tables, pictures, and figures from web documents such as those on confluence. What solutions can I use to extract and link figures in PDFs so that they are linked together in QnA bot requests?