we leverage batch write optimization which uses Copy-on-write B-tree variant enables high-throughput batch writes without blocking concurrent reads. MVCC-based design eliminates lock contention and provides predictable write amplification.