When ingesting your data, Amazon Bedrock first splits your documents or content into manageable chunks for efficient data retrieval. The chunks are then converted to embeddings and written to a vector index (vector representation of the data), while maintaining a mapping to the original document. T…