San Francisco Bay Area
Engineer Manager with strong tech experience across product and infra. Collaborate with cross-functional partners (PM, DS, UXR) to drive roadmap and deliver products with strong business values. Engage Ops partner to launch improvements to reduce operational cost. Continuously lead team to scale up systems and reduce risks. Strong expertise in infra and proven history of developing large-scale (millions QPS) distributed services.
Leading an eng team of 10+, across frontend and backend. Team scope includes Crypto trading, restrictions, settlement, accounting, tax, statements, etc. To highlight a few deliveries: Launched several powerful products: - Listed 8 new Crypto assets, including USDC in 2022, which was #1 ask from customers. In the meantime, reducing eng efforts for listing new assets from months to a couple weeks, reducing # of eng team involved from 10+ to 3. - Launched an epic new feature, Learn & Earn, to the platform. The first provider AVAX was launched in Dec 2022. More than 1M users accessed it within a few months. The platform was built in a extensible way that we launched a second provider USDC in just one month. - Launched Crypto statements that deliver to millions of Crypto users monthly. Scaled up Crypto trading services, which was one of the top company risks in 2021. - As IC & TL, launched the sharding project to make Crypto horizontally scalable. - As EM, invest heavily in building observability through setting up loadtest env. Achieved 10x improvements in both read and write path.
Led a team of 5 people to make the infra of Crypto trading service shardable and setup the second shard in production. This offload 90% of new user traffic from the original shard, and make Crypto trading service horizontally scalable for any potential traffic growth in the future.
Part of Service Discovery and Communication team, whose domain includes DNS, gRPC, load balancer, etc.
Major founding engineer and tech lead of a team of 5 people to provide real-time Materialized View on Spanner. - Designed the high level system, entire service workflow (including user registration, config management, hotspot handling (>1k per-key QPS)) and global server placement strategy. - Led the team to land the service to a smooth integration and production. Launched first few clients successfully with data freshness <2s at 99th with 0 outage. Clients include Youtube Content Management, Google Guice dependency information display, etc. Owned several critical areas of a real-time aggregation system, which powers Google Drive, Payment, Youtube, Map, etc, with tens of millions of QPS. - Domain expert of the write path, did several major changes (such as cache) to reduce write path latency and improve reliability. - Other owned features include providing read-after-write consistency from an eventually-consistent system, using aggregate results as index ranking signals, etc. - Led production excellence efforts by overseeing the reliability and robustness of the system, and collaborating with SRE and developers to resolve production issues.
Owner of L3 protocol (Distributed Logical Router support) for NSX-6.x and NSX-transformers Control Plane. Also an expert on overall control plane architecture, including cluster, CLI, and other components. Filed 4 patents.