An agent's document embeddings and query embeddings come from incompatible model versions. Which approach best fits?

1. Choose and evaluate an appropriate chunking strategy with metadata.
2. Align and version the embedding representation on both sides.
3. Preserve source provenance and validate answer support against retrieved evidence.
4. Evaluate reranking after candidate retrieval.