Senior Applied Scientist, Efficient LLM Inference & Model Optimization
Core
Turn frontier LLM and VLM inference bottlenecks into research questions, conduct rigorous experiments, and translate findings into production capabilities for the Nebius Token Factory.
Role type
Senior Applied Scientist (Efficient LLM Inference & Model Optimization)
Builds
Production inference engines, distributed inference architectures, and optimized serving systems for LLMs and VLMs.
Domain
AI/ML Systems, Large Language Models, Distributed Computing, Cloud Infrastructure
Deliverable
production ML models
Required skills
LLM/VLM inference, quantization, distillation, speculative decoding, KV-cache optimization, model/runtime co-optimization, distributed inference, experimental design, PyTorch, CUDA, Triton
Preferred skills
vLLM, SGLang, TensorRT-LLM, FlashAttention, FlashInfer, MoE routing, PDD, WideEP, RLHF, synthetic data generation
Technologies
PyTorch, Triton, CUDA, vLLM, SGLang, TensorRT-LLM, FlashAttention, FlashInfer
Responsibilities
Lead research projects in efficient LLM/VLM inference from hypotheses to production handoff; Develop and evaluate methods spanning quantization, distillation, and speculative decoding; Build prototypes using PyTorch, Triton, or CUDA-adjacent tooling; Research LLM request routing and scheduling strategies; Investigate multi-node inference for dense and MoE models; Develop rigorous evaluations covering quality, latency, throughput, and cost; Collaborate with engineering teams to select research priorities; Share results through reports, blogs, and papers.
Seniority
Senior, hands-on IC with research leadership
