TL;DR

Research Engineer/Scientist (AI): Building learning and evaluation foundations for personalized, multimodal AI systems with an accent on RLHF, reward modeling, and preference-learning pipelines. Focus on designing frameworks for context-aware and adaptive model behavior that improves through user feedback over long-term horizons.

Location: Must be based in San Francisco, CA (Hybrid: 4 days/week in-office). Relocation assistance provided.

Salary: $380,000 – $445,000 + Equity

Company

OpenAI is an AI research and deployment company dedicated to ensuring that general-purpose artificial intelligence benefits all of humanity.

What you will do

  • Develop RLHF and post-training methods for multimodal AI models.
  • Build reward models and preference-learning pipelines to improve adaptive model behavior.
  • Design evaluation frameworks and rubrics that capture long-term user value and contextual appropriateness.
  • Experiment with policy improvement strategies using explicit feedback and model-based grading.
  • Collaborate with safety researchers to ensure personalization remains interpretable and bounded.
  • Prototype training recipes and data pipelines for product-relevant AI behaviors.

Requirements

  • Strong background in machine learning research with focus on RLHF, reward modeling, or post-training.
  • Experience with reinforcement learning, ranking, personalization, or human-in-the-loop evaluation.
  • Ability to design rigorous empirical experiments and reliable evaluation metrics.
  • Comfort working across the full stack from data generation to training runs and analysis.
  • Must be located in or willing to relocate to San Francisco, CA.
  • Ability to thrive in a cross-functional team environment with engineers, designers, and safety researchers.

Culture & Benefits

  • Cutting-edge work on frontier AI systems with significant real-world product impact.
  • Collaborative culture valuing diverse perspectives and human-centric AI development.
  • Commitment to safety, ethical AI, and long-term user benefit.
  • Competitive compensation package including significant equity.
  • Supportive environment providing reasonable accommodations for disabilities.