At Cheetah Digital, I define SLOs and error budgets for high-traffic services and built the Grafana dashboards used in monthly reliability reviews. I redesigned alert routing and wrote runbooks, helping cut mean time to resolve from 4.6 hours to 1.5 hours.
I introduced blameless postmortems and follow-up tracking, and built a change-management process that reduced change-caused incidents by 60% year over year. I also serve as incident commander for the platform and handle client-facing incident communications and RCA documents.
At SnapFinance, I built the Terraform infrastructure foundation and Jenkins pipelines as code. New environments went from two weeks to under two hours, and I standardized secrets management across repositories.
At Provectus, I built and maintained data pipelines for PepsiCo’s ad-placement prediction models using Python, Airflow, dbt, and Snowflake. Earlier, at BV Media and Experian Marketing Services / Cheetah Digital, I worked on AWS infrastructure, monitoring, client provisioning, and production readiness.
