Using Redis Semantic Cache for Dynamic Question-Answer Retrieval from PDF Documents

Question

I'm currently working on implementing Redis semantic caching to store question-answer pairs extracted from PDF documents. However, I'm facing a challenge: ensuring that the cache doesn't serve answers based on outdated content if the document undergoes updates. Can someone provide guidance on how to effectively utilize Semantic cache to handle this issue? Alternatively, if caching isn't the best solution here, I'm open to suggestions for alternative strategies to address this scenario.