Senior Machine Learning Engineer, AI Safety

What you'll do

• Develop datasets and specialized models to benchmark safety in LLM security, agentic safety, and content safety.
• Create training recipes and datasets for filtering data, RL environments, and teacher models across core safety tracks.
• Research and deploy new techniques for risk detection and instruction hierarchy beyond standard post-training.
• Evaluate and benchmark end-to-end systems for hallucinations, ML fairness, and multi-turn conversational use cases.
• Implement post-training techniques including Supervised Fine-Tuning (SFT) and Reinforcement Learning (RLHF/RLAIF).
• Address specific risks including backdoors, data poisoning, model deception, and autonomous agent execution risks.
• Scale safety solutions for multi-modal foundational models and large multimodal datasets.

What we're looking for

• Master’s or PhD in Computer Science, Electrical Engineering, or a related quantitative field (or equivalent experience).
• 8+ years of proven experience in systems software engineering or machine learning engineering.
• 4+ years of hands-on work experience in post-training of LLMs, including SFT, RLHF/RLAIF, safety data generation, and production deployment.
• 1+ years of dedicated experience or research in LLM Security, Frontier Risks, Agentic Safety, or Multi-turn Safety Evaluation.
• In-depth knowledge of machine learning principles and frameworks (PyTorch preferred) with strong Python programming skills.
• Experience working with large multimodal datasets and multi-modal foundational models.
• Published papers as a primary author at top-tier conferences on AI Safety, alignment, or machine learning security (preferred).
• Proven experience with alignment/fine-tuning of Vision-Language Models (VLMs) or any-to-text foundational models (preferred).