Defend the defender. Hunt at scale.

Role overview Improve and defend a security platform whose detection decisions are made by statistical models rather than language models. The role combines detection engineering, adversarial testing, threat research, and technical participation in open standards for audit, agent identity, and revocation.

Responsibilities - Tune and evaluate four production detection approaches covering anomalies, sequences, drift, and cross-dimensional correlation; investigate false positives and emerging attack surfaces. - Red-team platform components, including its proxy, execution controls, and emergency stop mechanisms, to find weaknesses before adversaries do. - Conduct research on real threats and use findings to improve the platform’s threat-intelligence capabilities. - Coordinate technical correspondence on standards related to audit formats, agent identity, and revocation. - Explain detection choices and the distinction between useful signals and noisy alerts to technical and nontechnical audiences.

Requirements - Demonstrated experience with detection at scale, a substantial defensive or offensive security team, or threat research that has informed other practitioners. - Ability to interpret packet captures and proxy logs together to reconstruct activity. - Sound judgment about security decisions and the risks of using language models in detection or enforcement. - Experience producing work that engages regulatory or compliance concerns.

Nice to have - CVE attribution or published, cited threat research. - Experience with standards bodies such as IETF, W3C, NIST, or ISO. - Ability to assess and defend scoping security controls around AI action types.