Senior Go backend / platform engineer. Distributed systems at Chinese SaaS scale — 30k tenants, 6M live QR codes, 200K delayed jobs per second.
Stack: Go · Kafka · Redis · MongoDB · TiDB CDC · Flink · Kubernetes
Open source — hourglass: delayed jobs & cron for Go. SQLite embedded, sharded MySQL/TiDB + etcd cluster, same engine — time-wheel precision, SQL truth, epoch-fenced failover. Clean-room rebuild of a 200K jobs/s production scheduler (write-up). Also contributing to Apache pulsar-client-go (encryption & batch deserialization, consumer lifecycle), arrow-go (Parquet modular encryption, pyarrow interop), and dubbo-go / gost (global logger data race).
Recent writing — full archive at edwardshen125.github.io:
- One Queue for Every State Change: Fixing Silent Over-Assignment in a Live QR Code System — serializing WeCom QR-code membership updates through one Kafka topic, against the vendor's rate limits
- Running a Million-DAU Casual Game on Go: Rooms as the Concurrency Boundary
- Debugging a Sporadic Go OOM in a Recursive JSON Encode Path
Currently open to senior backend / platform roles, remote or relocation.


