AI Training Infrastructure Engineer, Perplexity AI

• Proven experience with large-scale LLMs frameworks building
• Strong in Python/Pytorch; C++/CUDA is a plus
• Self-starter with a willingness to take ownership of tasks
• Passion for tackling challenging problems
• Minimum of 6 years of working on relevant projects

Desirable

• Perplexity is seeking experienced AI Research Engineers and Scientists to continue to improve our in house Online LLMs, the Sonar models
• Your job is to work with team and create a robust and effective training framework (on top of Megatron/PyTorch), especially for post training LLMs
• Build a post training framework that can run cutting-edge model training jobs in scale
• Implement the necessary infra and components to support latest models and algorithms like SFT, RL (DPO/GRPO) and more
• Own the full stack data, training, and eval pipelines required to post-train LLM models
• Work closing with engineering teams to integrate Sonar models into our product