Describe how you would build an ingestion pipeline that feeds structured data into Elasticsearch for low-latency search and query modeling at Kharon.
