🛠 Tech Stack

💼 About This Role

You'll research data strategies for post-training LLMs, blending human feedback and synthetic data. Your core impact is designing scalable pipelines for high-quality data and improving model alignment. You'll work at a fast-moving AI lab with access to frontier-scale compute and data.

🎯 What You'll Do

• Design and execute data collection and synthesis strategies for post-training.
• Develop scalable pipelines for human and synthetic data generation.
• Research and model human preferences to improve reasoning and truthfulness.
• Design evaluation metrics and benchmarks for data quality and alignment.

📋 Requirements

• Strong engineering skills with ability to contribute code in complex codebases.
• Experience with data curation or synthetic data generation for LLMs.
• Proficiency in Python and at least one deep learning framework (PyTorch, TensorFlow).
• Strong grasp of probability, statistics, and ML fundamentals.

✨ Nice to Have

• Prior experience with RLHF, RLAIF, or reward learning for large models.
• Experience managing human data collection campaigns or annotation workflows.
• Familiarity with synthetic data pipelines or active learning.

🎁 Benefits & Perks

• 🚀 Fast-moving environment with minimal process overhead.
• 💻 Access to frontier-scale compute and data from day one.
• 🏆 Small, highly selective team where research reaches deployment quickly.