LLM Post-Training Engineer Graduate (Research & Product) - 2027 Start
Core
Building, researching, and applying Large Language Models (LLMs) to power global multimedia products, specifically focusing on post-training strategies and model alignment.
Role type
Graduate LLM Post-Training Engineer (Research & Product)
Builds
Production-ready AI applications for video understanding, translation, and content classification
Domain
AI / Machine Learning / Multimedia
Deliverable
production ML models
Required skills
Python, PyTorch, JAX, Transformer architectures, LLM training principles, large-scale dataset analysis
Preferred skills
LLM fine-tuning, RLHF, synthetic data generation, distributed training tools (DeepSpeed, Megatron-LM), multimodal AI
Responsibilities
Support development and optimization of post-training strategies (instruction tuning, SFT/DPO/PPO), build evaluation pipelines for model performance and safety, implement reward modeling and human preference learning, collaborate with engineering teams to bridge research and production, analyze large-scale datasets to improve model behavior.
Seniority
Graduate (Master's level)