Principal Research Engineer
Core
Lead model-training programs including pretraining, fine-tuning, and post-training to improve model quality, safety, and performance.
Role type
Principal Research Engineer (Large Language Models)
Builds
Foundation models and their adaptations via pretraining, fine-tuning, and post-training pipelines.
Domain
Artificial Intelligence / Large Language Models
Deliverable
production ML models
Required skills
model pretraining, supervised fine-tuning, post-training, model architecture design, training data curation, experiment design, distributed training, hyperparameter optimization, debugging training instability, responsible AI practices
Preferred skills
reinforcement learning, model distillation, parameter-efficient fine-tuning, synthetic data generation
Technologies
PyTorch, JAX, Python, C, C++, C#, Java, GPU clusters
Responsibilities
Design and implement controlled experiments and evaluation methods for accuracy, reasoning, and safety; Diagnose and resolve training instability, convergence failures, and data-quality defects; Partner with infrastructure specialists to scale training across GPU clusters; Provide technical direction and mentor engineers across research and engineering teams.
Seniority
Principal, strategy & mentorship
