SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Nebius is a Nasdaq-listed AI cloud infrastructure company headquartered in Amsterdam with R&D hubs across Europe, North America, and Israel. The company provides a full-stack AI cloud platform supporting developers and enterprises from data and model training through production deployment.
Nebius Token Factory is seeking a Senior Applied Scientist to lead research projects in efficient LLM and VLM inference. You will own focused research and optimization initiatives, turning frontier inference bottlenecks into rigorous research questions, conducting experiments, and translating findings into production capabilities.
Key responsibilities include:
- Leading research projects spanning quantization, distillation, speculative decoding, KV-cache optimization, and model/runtime co-optimization
- Developing and evaluating methods including quantization-aware training (QAT), long-context inference, MoE routing, and distributed inference architectures
- Building prototypes using PyTorch, Triton, CUDA, and inference-serving frameworks, collaborating with MLEs and platform engineers to transition them to production
- Researching LLM request routing and scheduling strategies, including cache-aware load balancing and prefill–decode disaggregation (PDD)
- Investigating multi-node inference for dense and mixture-of-experts models, including wide expert parallelism (WideEP)
- Developing rigorous evaluations covering quality, latency, throughput, numerical stability, memory footprint, tail latency, and cost per token
- Working cross-functionally with MLE, GPU kernel, backend infrastructure, product, and customer teams to prioritize research with measurable production impact
- Sharing results through internal reports, technical blogs, papers, and open-source artifacts; mentoring engineers and scientists on experimental design and model/system trade-offs
Requirements:
- PhD in computer science, machine learning, ML systems, computer systems, computer architecture, electrical engineering, applied mathematics, or closely related discipline
- Strong publication record or equivalent research artifacts in ML, ML systems, efficient inference, model compression, quantization, distillation, serving systems, or related areas
- Strong Python and PyTorch implementation skills with ability to turn ideas into experiments and working prototypes
- Deep knowledge of LLMs, VLMs, transformer inference, decoding algorithms, model compression, quantization, and production-serving trade-offs
- Strong experimental design skills covering ablations, baselines, metrics, statistical reasoning, and failure analysis
- Excellent written and verbal communication
Nice-to-haves:
- First-author publications at top venues (NeurIPS, ICML, ICLR, MLSys, ACL, EMNLP, ASPLOS, OSDI, SOSP, ISCA, HPCA)
- Production ML model deployment or inference optimization experience
- Experience with vLLM, SGLang, TensorRT-LLM, NVIDIA Dynamo, FlashAttention, FlashInfer, Triton, CUDA, or PyTorch internals
- Experience with post-training, SFT, DPO, RLHF, RLAIF, preference optimization, or synthetic data generation
- Open-source research artifacts, benchmarks, technical blogs, or invited talks demonstrating contributions to efficient AI systems
- Research or implementation experience in distributed inference system architecture (LLM request routing, PDD, multi-node serving, MoE expert parallelism, WideEP)
