MIGHT User

I’m a Staff Site Reliability Engineer / Platform Engineer with 12+ years building and operating large-scale cloud infrastructure across TikTok, Snap, and Amazon. I specialize in Kubernetes reliability, infrastructure automation, and observability—always with the goal of improving uptime and speeding up safe deployments.

At TikTok, I built Flink-based aggregation and anomaly-detection pipelines and used Prometheus alerts to reduce regression detection time from ~20 minutes to under 5 minutes. I helped resolve ranking latency incidents with canary rollbacks and strengthened operational readiness through runbooks, RCA templates, and alert-routing guidance. I also improved service resilience with retries, timeouts, circuit-breakers, and fallbacks, and drove adoption of SLO dashboards and runbook patterns across sub-teams.

At Snap, I owned SLOs for Camera and AR platform services and built release-safety automation that combined CI/CD checks, synthetic validation, and staged rollout gates. I migrated AR backend services to ECS/EKS to support zero-downtime rollout practices, and I improved monitoring by creating Grafana dashboards and PagerDuty alerts to reduce detection time from hours to minutes.

Earlier in my career, I helped found Viro Media’s platform by implementing AWS backend infrastructure, Terraform-based self-service deployments, and scalable streaming services. My mindset stays the same: automate guardrails, measure outcomes with SLOs, and make incidents easier to detect, diagnose, and recover from.