LLM Post-Training Engineer Graduate (Research & Product) - 2027 Start

What you'll do

• Develop and optimize instruction tuning, preference tuning, and model alignment strategies.
• Build evaluation pipelines and participate in reward modeling and human preference learning research.
• Analyze large-scale datasets and collaborate on production AI applications such as video understanding, translation, and content classification.

What they ask for

• Completing or recently completed a Master's degree in Computer Science, Machine Learning, or a related discipline.
• Strong Python programming skills and experience with PyTorch or JAX.
• Foundational understanding of Transformer architectures and LLM training principles.
• Ability to learn quickly in a fast-paced environment and strong interest in applying AI to products.

Nice to have

• Experience or research involving LLM fine-tuning, RLHF, or synthetic data generation; familiarity with DeepSpeed or Megatron-LM.
• Experience or interest in multimodal AI, or a track record of building and deploying AI projects.

Tools

Python, PyTorch, JAX, DeepSpeed, Megatron-LM, SFT, DPO, PPO

AI focus

LLM post-training and alignment, including instruction and preference tuning, reward modeling, human preference learning, evaluation, and multimedia product applications.