Senior Research Engineer (AI Safety)

Location: Remote internationally or in person in Berkeley, California, or Singapore; remote hiring is available in most countries.

Salary: $225,000–$350,000 per year, adjusted for experience and location.

Company

hirify.global is a non-profit AI research institute focused on AI safety, AI risk research, and coordinated responses to advanced AI risks.

What you will do

• Lead projects in detecting and preventing deceptive behavior in advanced AI models.
• Apply research insights to detect and mitigate vulnerabilities, catastrophic misuse, and other risks in frontier models.
• Build research frameworks and infrastructure that enable larger experiments and faster iteration.
• Train transformers and conduct adversarial evaluations, red-teaming, and robustness research.
• Develop high-performance distributed systems, GPU infrastructure, and tooling for fine-tuning and post-training workflows.
• Join existing workstreams and potentially contribute to new areas such as mechanistic interpretability.

Requirements

• Significant software engineering experience, demonstrated through prior work or open-source contributions.
• Fluency in Python.
• Results-oriented approach and motivation to conduct impactful research.
• Prior experience mentoring engineers or scientists in engineering skills.
• Expertise in at least one core area: machine learning, high-performance computing, or technical leadership.
• For machine learning: substantial experience training transformers with frameworks such as PyTorch or JAX, plus knowledge of linear algebra, calculus, vector probability, and statistics.
• For high-performance computing: experience with Kubernetes or SLURM, high-performance distributed systems, and code optimization and profiling, ideally including GPU or CUDA work.
• For technical leadership: experience designing large-scale software systems or leading major refactors, plus project management experience with small teams and 3–6 month research roadmaps.

Culture & Benefits

• Full-time schedule of 40 hours per week.
• USA- and Singapore-based employees are employed directly; candidates outside those countries are engaged through an EOR organization or as contractors.
• Visa sponsorship is available for in-person employees.
• Work-related travel and equipment expenses are covered.
• Catered lunch and dinner are available at the Berkeley office.

Hiring process

• Complete a 72-minute programming assessment.
• Participate in two interviews with technical staff.
• Complete a 1–2 week paid work trial; alternative assessment arrangements may be available if a work trial is not possible.
• Apply through the application form; emailing a resume directly does not affect review.