I
Product Manager, Agentic AI Security Environments
Sin especificarTiempo completoGestión de productoAI Services
Own the product line for agentic AI security environments, working across research, safety, engineering, and commercial teams to turn customer needs into products labs want to use and buy.
Responsabilidades
- Own product strategy, roadmap, and catalog across cyber misuse, misalignment testing, and agentic safety environments.
- Work with safety, alignment, and reinforcement learning teams at frontier labs to understand their needs and gaps in existing environments.
- Write specifications covering scenarios, agent tools and actions, success criteria, and quality standards.
- Set and track metrics for realism, difficulty, grading reliability, and resistance to reward hacking and evaluation awareness.
- Work with leadership and sales on packaging and pricing, including licensed libraries, subscriptions, and custom builds.
- Scope client engagements, manage expectations, and feed client feedback into the roadmap.
- Track the reinforcement learning environment and AI security landscape, competitors, and research.
Requisitos
- At least 5 years of product management experience with technical products such as security, developer tools, machine learning infrastructure, or data and evaluation platforms.
- Strong understanding of offensive security, including attack chains and realistic security scenarios.
- Hands-on understanding of AI agents and training, including tool use, agent harnesses, reinforcement learning environments, rewards, verifiers, and reward hacking.
- Experience selling to or partnering with technical buyers, such as research, security, or engineering teams.
- Strong writing skills for specifications and roadmaps, and sound product judgment.
- Comfort with product packaging, pricing, and product-line economics.
- Ability to work effectively amid ambiguity and align multiple stakeholders.
Se valora
- Prior experience as an engineer, penetration tester, or security researcher.
- Product experience at an AI data, evaluation, or reinforcement learning environment company, or a cyber range or security training platform.
- Direct experience working with frontier AI labs.
- Familiarity with AI safety and alignment research, including misalignment, evaluation awareness, and agentic risk.