- Build and operate Claude Code environments, skills, subagents, MCP servers, and guardrails.
- Define human checkpoints and approval gates for agent actions in production.
- Run and harden the AWS environment for availability, performance, cost, and security.
- Lead on-call incident response, root cause analysis, blameless postmortems, and runbook development; automate recurring fixes.
- Own infrastructure as code using Terraform and some CloudFormation.
- Build observability for systems and agents using Datadog, Grafana, and PagerDuty, including alerting, tracing, and token and cost tracking.
- Maintain shared AWS services, including MSK, RDS, networking, and DNS records.
- Build reusable Terraform modules, CI/CD actions, and base container images.
- Triage infrastructure requests and provide hands-on support to product development, platform, and support teams.
- Document CI/CD and IaC patterns, run lunch-and-learns, and identify cloud and LLM cost savings.