If you decide to not stream enrichment data into the enrichment store, you can bulk load it from HDFS, which involves configuring an extractor file, configuring element-to-enrichment mapping, running the loader, and mapping fields to Apache HBase enrichments.