Applied Scientist: Efficient LLM Inference & Optimization

Nebius Group; United States; Remote; USD 150,000 - 190,000; Full time

Benefits offered by this job

Competitive compensation

Career growth opportunities

International environment

Job summary

Nebius Group is seeking a leading ML research scientist to advance efficient LLM and VLM inference across end-to-end production. You will own focused research and translate findings into scalable components for data and model workflows.

You will develop and evaluate methods spanning quantization, distillation, speculative decoding, KV-cache optimization, and model/runtime co-optimization, building prototypes with PyTorch and Triton and collaborating with engineers to ship production-quality

Qualifications

• PhD in computer science, ML systems, ML or related field.
• Strong publication record in ML, ML systems, efficient inference, model compression or related areas.
• Strong Python and PyTorch implementation skills; ability to turn ideas into experiments and prototypes.
• Deep knowledge of LLMs, VLMs, transformer inference, decoding, model compression and production-serving trade-offs.
• Excellent written and verbal communication.

Responsibilities

• Lead research projects in efficient LLM and VLM inference from hypotheses to production handoff.
• Develop and evaluate methods spanning quantization, QAT, distillation, speculative decoding, KV-cache reuse/compression, long-context inference, MoE routing, and model/runtime co-optimization.
• Build prototypes using PyTorch, Triton, CUDA-adjacent tooling, or inference-serving frameworks; collaborate with MLEs and platform engineers to productionize.
• Research LLM request routing and scheduling strategies, including cache-aware load balancing and prefill-decode disaggregation (PDD).
• Share results through internal reports, technical blogs, papers, and open-source artifacts; mentor engineers and scientists on experimental design and trade-offs.

Skills

Python

PyTorch

LLMs

VLMs

Transformer inference

Quantization

Distillation

QAT

MoE routing

KV-cache

Education

PhD in CS/ML systems/related

Tools

PyTorch

Triton

CUDA

TensorRT-LLM

Job description

Nebius Group is seeking a leading ML research scientist to advance efficient LLM and VLM inference across end-to-end production. You will own focused research and translate findings into scalable components for data and model workflows.

You will develop and evaluate methods spanning quantization, distillation, speculative decoding, KV-cache optimization, and model/runtime co-optimization, building prototypes with PyTorch and Triton and collaborating with engineers to ship production-quality