Member of Technical Staff, AI Red Teaming

Safety data needs to keep pace with capabilities. As AI advances, we build evaluations that measure consequential agent failures and produce training data that helps mitigate them.

Although we're only a year old, we've already been featured in Anthropic's Fable 5 & Mythos 5 System Card, led Anthropic's prompt injection evaluation of Auto mode, and been featured in Meta's Muse Spark Safety & Preparedness Report.

Our areas of focus include red teaming, control, anti-distillation, AI safety reasoning, reward-hacking propensity evaluations, and embedded assessments.

A lot of companies say they use AI extensively. We mean it. PR count isn't the goal, but to give a concrete example, last week a single engineer merged 1,000 PRs.

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

We're looking for an engineer who uses coding agents extensively and wants to grow through hands‑on technical work: building, testing, and debugging systems. Technical leads guide project direction and lead customer relationships, while you take increasing ownership of the engineering work.

This is an early‑stage startup role, so you should expect and enjoy being in a role where responsibilities can grow and priorities can quickly change.

You'll work with our technical leads on difficult safety and alignment problems, building evaluations, RL environments, and training‑data pipelines. You'll write production code, develop tests, debug failures, and improve reliability, with support on architecture and project scoping.

You'll build the software factory for automated red teaming, from reusable attack tooling and agent workflows to the automated tests that make each new engagement faster to run and easier to trust.

You'll collaborate with our engineers and red teamers, building tools that make their work faster and taking on larger technical problems as your experience grows. The failures you find become evaluations and training data that labs use to make their models more robust.

• Impact-driven
  : you want to improve the safety of frontier AI systems.

• Agent-native engineering
  : you use coding agents and LLMs as core development tools.

• Strong software engineering
  : you can design, build, test, and debug substantial systems, including code produced by agents.

• Adversarial mindset
  : you enjoy finding the ways systems break, and turning one‑off exploits into repeatable, measurable attacks.

• Autonomy and judgment
  : you can break down an engineering task, make good trade‑offs, and keep the work moving, asking for guidance when you need it.

• Early‑stage startup drive
  : you enjoy a fast pace, shifting priorities, limited structure, and taking on whatever is most important to help the company succeed.

• Adversarial machine learning
  : experience building automated attacks on models, or research on robustness. This is especially valuable for this role.

• Successful jailbreaks or prompt injections against frontier models

• Experience in security or ML security

• Experience building evals, red‑teaming systems, RL environments, or model‑training data

• A strong mathematical or research background

These criteria are a guide, not a checklist. If you want to do your life's work making frontier models safer and this role excites you, we encourage you to apply.

Total compensation: Includes base salary, bonus, and the estimated annualized value of equity. Your level depends on the scope, autonomy, and impact we expect you to have:

• L4: $350,000

• L5: $520,000

• L6:…