Principal Research Engineer, Post-Training

What you'll do

• Set the technical roadmap and mentor research and engineering teams working on mid- and post-training.
• Lead development of alignment methods, training objectives, and evaluation frameworks to improve model behavior and performance.
• Design scalable training, data, inference, and serving systems, including optimization of distributed training and GPU utilization.

What they ask for

• PhD in Computer Science, Machine Learning, AI, or a related field, or equivalent industry experience.
• Significant experience leading machine learning, AI research, or large-scale distributed systems projects or teams, including mentoring teams.
• Deep understanding of transformers, reinforcement learning, alignment methods, and large language models; experience leading LLM training, mid-training, or post-training teams.
• Track record delivering production ML research or systems, with experience building production-quality ML infrastructure and training, serving, debugging, and optimizing large-scale GPU-based models.

Nice to have

• Hands-on experience adapting open-source models such as Mistral and Qwen for personas, creative writing, or role-playing.
• Familiarity with Kubernetes, Docker, and modern orchestration platforms; publications at leading ML conferences or AI community contributions.

Pay

$275K – $400K (base salary; currency not stated)

Tools

Mistral, Qwen, Kubernetes, Docker

AI focus

Post-trains large language models using reinforcement learning, preference optimization, supervised fine-tuning, and alignment approaches for engaging, aligned role-playing chat; develops evaluation and scalable training and inference systems.