AI Safety Engineer - Red Teaming & Evaluations

Trajectory Labs; Berkeley (CA); Hybrid; USD 250,000 - 400,000; Full time

Benefits offered by this job

Health coverage

401(k)

Catered meals

Moving allowance

Job summary

Trajectory Labs in Berkeley, CA is hiring a Member of Technical Staff to build the systems behind our red teaming of the world's most powerful AI models. You will work with technical leads to build attack agents, evaluation harnesses, and training-data pipelines that help make these models safer.

You will write production code, develop tests, debug failures, and improve reliability, with architecture guidance. This is an early-stage startup role where responsibilities grow with the company.

Qualifications

• Experience with AI safety or ML security preferred.
• Experience building evals, red-teaming systems, RL environments, or model-training data is a plus.
• Strong software engineering skills with ability to design, build, test, and debug large systems.

Responsibilities

• Build evaluations, RL environments, and training-data pipelines.
• Write production code, develop tests, debug failures and improve reliability.
• Create the software factory for automated red teaming and scalable engagements.
• Collaborate with engineers and red teamers to tackle large technical problems and produce actionable evaluations.

Skills

Impact-driven

Agent-native engineering

Strong software engineering

Adversarial mindset

Autonomy and judgment

Early-stage startup drive

Adversarial machine learning

Jailbreaks / prompt injections

Security / ML security

Evals / red-teaming / RL envs

Mathematical / research background

Job description

Trajectory Labs in Berkeley, CA is hiring a Member of Technical Staff to build the systems behind our red teaming of the world's most powerful AI models. You will work with technical leads to build attack agents, evaluation harnesses, and training-data pipelines that help make these models safer.

You will write production code, develop tests, debug failures, and improve reliability, with architecture guidance. This is an early-stage startup role where responsibilities grow with the company.