• Played an integral role in Meta’s Global Resource Management initiative. Owning the systems such as the Global Reservations Service, which enable region-agnostic resource allocation across Meta’s multiple data centers. This work simplified capacity provisioning, improved disaster-readiness by automatically redistributing resources, and ensured global optimization of service performance.

• Working in the Core Compute Infrastructure team, building a highly scalable resource management and allocation system that forms the critical backbone of Meta’s private Infra Cloud. This system efficiently manages a shared pool of millions of servers, providing hundreds of internal teams with a user-friendly interface to deliver reliable, cost-effective resources.

• Contributed to the AI computation capacity management initiative to automate GPU capacity operation and reduce manual churns. This project improved overall capacity efficiency, achieving significant savings, and directly contributed to optimizing Meta’s resource utilization.

• Designed and implemented a proxy-based distributed data synchronization protocol for NSX multi-site federation, enabling seamless data flow across multiple sites.

• Led the optimization and scaling projects of the Central Controller Module for VMware NSX using Java, ensuring robust performance under high workloads.

• Guided and mentored junior team members, providing technical guidance and facilitating their professional growth within the organization

• Innovated on the VMware NSX, focusing on the Central Control Plane.

• Transitioned a graph-based algorithm for distributed firewall policies from Java to Ddlog and Rust, doubling performance and memory efficiency.

• Played a key role in amplifying the Central Controller's support for the Distributed Firewall of VMware NSX.

• Revolutionized the data-synchronization framework by migrating from a single-threaded system to a multi-threaded producer-consumer model.