- Carry a normal on-call rotation and act as a hands-on responder for incidents.
- Investigate, fix, and document incidents including debugging application code and shipping fixes directly into service repos.
- Design, build, and ship LLM-based agents that integrate with Kubernetes, cloud APIs, and observability stacks.
- Define tool interfaces and build safe wrappers for agent interactions with production systems.
- Set guardrails for autonomous agent actions and manage human-in-the-loop workflows.
- Own agent evaluation by defining metrics and building test suites against historical incidents.
- Maintain security and compliance standards, including audit trails and least-privilege access.
AWSPythonGCP+5 more