Job Description
Step into the future with Nexus Horizon Systems, where we are architecting the ethical foundation of Artificial General Intelligence (AGI). We are looking for a talented Next-Gen AI Alignment Engineer to join our elite team in Seattle. As we prepare to launch our 2026 roadmap, your mission will be to ensure that advanced AI systems are not only powerful but also safe, interpretable, and aligned with human values.
In this high-impact role, you will work at the frontier of machine learning safety. You will design rigorous evaluation frameworks, implement state-of-the-art alignment techniques, and collaborate with world-class researchers to mitigate systemic risks. If you are passionate about the responsible development of AI and want to leave a lasting mark on the future of technology, this is your opportunity.
Responsibilities
- Develop and deploy advanced safety mechanisms for large language models and multi-agent systems.
- Conduct adversarial testing and red-teaming to identify and patch vulnerabilities before deployment.
- Design interpretability tools that make complex AI decision-making processes transparent and auditable.
- Collaborate with policy and ethics teams to define industry standards for AI governance.
- Research novel alignment theories, including constitutional AI and reinforcement learning from human feedback (RLHF).
- Mentor junior engineers and contribute to the broader safety research community.
Qualifications
- Master’s or PhD in Computer Science, Physics, Philosophy, or a related technical field.
- 5+ years of professional experience in Machine Learning, NLP, or Reinforcement Learning.
- Deep understanding of causal inference, game theory, or formal verification methods.
- Proficiency in Python, PyTorch, TensorFlow, and distributed computing frameworks.
- Strong track record of publishing in top-tier AI conferences (NeurIPS, ICML, ICLR) or open-sourcing significant safety tools.
- Exceptional problem-solving skills and ability to thrive in a fast-paced, ambiguous environment.