Worked on performance optimization for the MERGE INTO command in the apache/paimon repository, focusing on reducing execution time and resource usage during data ingestion. Developed a shortcut that maps source attributes directly to the target table’s _ROW_ID, effectively bypassing the need for full-join MERGE paths. This approach improved workload speed and lowered both CPU and memory consumption. The solution required deep integration with Spark, advanced SQL optimization, and low-level data modeling, all implemented in Scala. The work demonstrated strong data engineering and database management skills, resulting in faster MERGE operations and more efficient handling of large-scale data processing tasks.