Worked on the sei-protocol/sei-chain repository to deliver a performance-focused enhancement in the RPC path, specifically targeting the eth_getBlockByNumber and eth_getBlockByHash calls. Implemented an in-place Bloom filter calculation using Go, which eliminated the need for temporary slice allocations when combining block blooms. This approach reduced memory allocations and garbage collection pressure, resulting in improved latency and throughput for blocks containing multiple EVM or WASM transactions. The work involved backend development and close cross-team collaboration, including code review, to ensure robust integration. The feature addressed a critical performance bottleneck in high-traffic RPC endpoints, enhancing overall system efficiency.