Harmonic is building the world's most advanced mathematical reasoning engine and needs to scale its reinforcement learning systems beyond human-level capabilities.
Requirements
- 2+ years of experience focused on training large language models
- Knowledge of cutting-edge models, and experience building evaluations for model capability
- Expertise in Python and PyTorch
- Experience with distributed training, parallel computing, and GPU acceleration
- Strong experience with large-scale GPU clusters, HPC environments, and job scheduling/orchestration tools (e.g., SLURM or Kubernetes).
- Knowledge of reinforcement learning techniques
Responsibilities
- Conduct research and implement solutions in areas such as model architecture, algorithms, data processing, and optimization.
- Optimize and scale our training infrastructure to improve efficiency and reliability in a reinforcement learning setting.
Other
- BS in Computer Science, related technical fields, or equivalent industry experience
- MS or PhD in Computer Science, Mathematics, or related technical fields
- Contributions to open-source AI projects
- Unlimited PTO
- 401(k) matching
