AI Infra Engineer Graduate (Recommendation & LLM) 2027 Start (PhD)

What you'll do

• Build infrastructure for large-scale model training and online inference.
• Develop distributed systems and optimize GPU performance and communication for recommendation models and LLMs.
• Collaborate with researchers to deploy LLM training and serving solutions and address system bottlenecks.

What they ask for

• Completing or recently completed a PhD in Computer Science or a related discipline.
• Strong programming skills in C++ or Python.
• Good understanding of data structures, algorithms, and computer systems; familiarity with PyTorch or TensorFlow.
• Knowledge of Transformer architectures and LLMs.

Nice to have

• Hands-on LLM training or inference experience through research, internships, or open-source projects; familiarity with distributed training concepts.
• GPU programming experience with CUDA, Triton, or similar technologies; understanding of LLM serving techniques such as KV Cache, Continuous Batching, or FlashAttention.

Tools

C++, Python, PyTorch, TensorFlow, CUDA, Triton, FSDP, ZeRO

AI focus

Develop and optimize large-scale recommendation and LLM training, inference, serving, and GPU infrastructure.