Research Engineer (AI)

TL;DR

Research Engineer (AI): Advancing search and knowledge post-training to build trustworthy AI systems with an accent on reinforcement learning, retrieval, and evaluation. Focus on designing controlled search environments, building frontier-discriminating evaluations, and optimizing training rigor.

Location: Must be based in the US (San Francisco, Seattle, or New York City) and attend the office at least 25% of the time.

Salary: $500,000 - $850,000 USD per year

Company

AI safety and research company focused on creating reliable, interpretable, and steerable AI systems.

What you will do

• Own end-to-end research directions for search post-training problems, from hypothesis to training.
• Build instrumentation for controlled experiments to study how environment factors contribute to capabilities.
• Design frontier-discriminating evaluations to distinguish genuine reasoning from pattern matching.
• Drive optimization rigor across the stack, including efficient experiment design and training run economics.
• Collaborate with post-training, RL infrastructure, and product teams to translate model behavior into signals.
• Establish the team's experimental standards for measurement and result validation.

Requirements

• Outstanding software engineering skills in Python, spanning data pipelines to RL training.
• Proven track record of repeatedly shipping real ML research.
• Rigorous quantitative mindset, utilizing ablations, controls, and confidence intervals.
• Ability to operate with high autonomy in ambiguous environments.
• Strong written and verbal communication skills for defending design choices.
• Bachelor's degree or equivalent combination of education and experience.

Nice to have

• Hands-on experience with RL on large language models (reward design, scaling behavior).
• Background in search, retrieval, RAG, or agents that reason over external sources.
• Experience building evaluations for open-ended or knowledge-intensive LLM behavior.
• Prior work in frontier AI labs or quant research firms.
• Published research on LLMs, RL, retrieval, or calibration.

Culture & Benefits

• Collaborative "big science" approach focusing on high-impact research.
• Competitive compensation and optional equity donation matching.
• Generous vacation and parental leave policies.
• Flexible working hours.
• Visa sponsorship provided.