What you’ll do
- Architect runtime policy engines that govern agent tool-use, data access, and decision authority.
- Build detection and response primitives for prompt-injection, data-exfiltration, and rogue-agent scenarios.
- Partner with customer security teams during deployments and incident reviews.
- Drive the agent-security threat model used across our reference architecture.
What we’re looking for
- 5+ years in security engineering — application, cloud, or identity security.
- Hands-on with at least one LLM agent framework and the OWASP LLM Top 10.
- Strong systems-thinking; comfortable in Python or Go and modern cloud (AWS/Azure/GCP).
- Track record of shipping security software that customers actually deployed.
Nice to have
- Experience with policy engines (OPA, Cedar) or zero-trust architectures.
- Prior contributions to security frameworks or open-source security projects.