When using AI enrichment on blob storage, which feature helps to reduce costs related to search service?

Enrichment caching is a feature that significantly reduces costs related to search services when using AI enrichment on blob storage. This feature allows the results of enrichment processes, such as extracting content or performing analysis, to be temporarily stored. When the same data is accessed again, instead of incurring additional costs by re-enriching the data, the cached results can be retrieved quickly and efficiently. This is especially beneficial in scenarios where the same data is processed multiple times or when handling large datasets, leading to savings on both processing time and associated costs.

Other features like data compression, document batching, and service tier scaling, while useful in their own contexts, do not specifically target cost reduction in the same way. Data compression helps optimize storage but does not directly impact enrichment costs. Document batching focuses on processing multiple documents at once for efficiency but does not cache results from prior enrichment. Service tier scaling adjusts the performance levels of the service depending on usage needs but does not inherently reduce costs associated with repeated enrichment processing.