CareerPlanSign in

Principal Security Research Manager

United States, Washington, Redmond💼 Full-time🗓 2026-07-24 → 2026-09-25

Core

Design and build end-to-end evaluation harnesses for agentic security workflows, creating benchmark suites and integrating evaluations into engineering pipelines to ensure reproducible evidence before production rollout.

Role type

Principal Security Research Manager (Agentic Security Evaluation)

Builds

Evaluation harnesses, benchmark suites, golden datasets, automated grading pipelines, and dashboards for security agent workflows.

Domain

Cybersecurity, AI Safety, Agentic Systems, Software Engineering

Deliverable

production ML models | product features | dashboards & analysis

Required skills

software engineering, system design, evaluation harness implementation, dataset creation, statistical analysis, security engineering, vulnerability management, LLM evaluation, automated grading, CI/CD integration, people management

Preferred skills

threat analysis, anomaly detection, SAST/SCA, SARIF, secure development lifecycle, experimentation design, human-review protocols

Technologies

LLMs, AI agents, automation frameworks, data pipelines, APIs, engineering telemetry, structured data, logs, traces, metrics

Responsibilities

Design and implement deterministic validators and workflow adapters for security agent evaluations; Create representative benchmark suites including adversarial and production-derived cases; Integrate evaluation results with production feedback and detect regressions; Partner with cross-functional teams to translate evaluation results into release recommendations; Convert learnings into reusable playbooks, templates, and reference implementations.

Seniority

Principal, strategy & mentorship

Sourced via microsoft · Listed on CareerPlan, which tracks 70,000+ jobs from 20+ sources.