Salary
💰 $200,000 - $375,000 per year
About the role
- Design and implement post-training systems and methodologies in close partnership with research scientists and domain experts
- Build and maintain infrastructure that supports large-scale model training, specialized data processing, and benchmark evaluation
- Develop robust frameworks for verifying the quality and integrity of highly specialized domain datasets
- Create next-generation LLM benchmarks that push the boundaries of model evaluation and capabilities assessment
- Optimize performance across software and hardware layers to accelerate post-training experimentation and deployment
- Collaborate across disciplines to ensure rigorous validation of model improvements and benchmark reliability
Requirements
- Strong Python programming skills with attention to clean, efficient, and scalable code
- Experience building and operating large-scale systems for model post-training, specialized data processing, or benchmark evaluation
- Deep familiarity with PyTorch and modern post-training techniques (RLHF, constitutional AI, etc.)
- Background in applied machine learning, model evaluation, or large-scale data quality assessment
- Experience with benchmark design, evaluation methodologies, and performance measurement frameworks
- Clear communication skills and a collaborative mindset for cross-functional research teams
- Extra credit: Experience optimizing deep learning models for performance (memory, training speed)
- Extra credit: Interest in the societal and ethical impacts of AI technologies
- Extra credit: Contributions to open-source ML infrastructure or tools