ByteDance — Volcano Engine
Jun 2023 — PresentSenior Software Engineer 2-2 · Cloud & Kubernetes
- Cloud Auto-Remediation Framework — architected a configurable, extensible system for issue detection, policy definition and automated remediation; owned core components: API & models, anomaly ingestion, policy engine, remediation engine and reporting.
- LLM-based Cloud Assistant — built an AI assistant that answers questions via RAG, performs tasks like lookups and machine reboots, and runs diagnostics on cloud resources using LLM-generated scripts from runbooks and engineer corrections.
- Developer Experience for Diagnostics Platform — delivered tooling for efficient DSL development in TypeScript with full type info, code generation, custom compilation, local/remote debugging and LLM-based auto-i18n.
- Internal Diagnostics Platform — led API and architecture redesign; performance work, bugfixes and diagnostic scripts.
- Dynamic Configuration Service — overhauled API, architecture and performance; added cloud-native (IP→label) features, robust monitoring, and GitOps-based config management.
- K8S & ETCD Reliability — drove the "adaptive throttling" initiative, hardened ETCD backups, migrated large objects out of ETCD, and was key to APIServer incident recovery and post-mortems.