4

42dot

Unified platform for multi-modal urban mobility

LLM Engineer - Optimization

Applied Machine LearningUpdated on Nov 11
Full-timePangyo-ro, Bundang-gu, Seongnam-si, Gyeonggi-do, South KoreaNo salary listedHybrid

About the Job

Requirements

  • At least 3 years of experience in LLM, machine learning infrastructure, or inference optimization.
  • Experience developing an LLM inference engine or AI runtime.
  • Understanding of GPU architecture, CUDA programming, or parallel computing.
  • Understanding of model optimization techniques such as quantization, model compression, or compiler optimization.
  • Experience using deep learning frameworks such as PyTorch, ONNX, or TensorRT.
  • Proficiency in at least one of Python or C/C++, with software engineering capability.
  • Ability to analyze performance and solve problems.

Responsibilities

  • Optimize inference performance, including latency, throughput, and memory efficiency, for large language models.
  • Research and apply optimization techniques suited to different model architectures and inference environments.
  • Improve performance in real-world service scenarios, including long-context and multi-turn conversations.
  • Develop and optimize GPU- and accelerator-based LLM inference engines.
  • Use or improve inference frameworks such as vLLM, TensorRT-LLM, SGLang, llama.cpp, ONNX Runtime, and MLX.
  • Apply current serving techniques such as speculative decoding and prefill-decode disaggregation.
  • Research and apply model-compression methods, including quantization, pruning, and distillation.
  • Use compiler and kernel optimization technologies such as CUDA, Triton, TensorRT, TVM, and MLIR to improve inference performance.
  • Analyze and optimize GPU memory and kernel efficiency.
  • Develop techniques to run LLMs efficiently on mobile, embedded, and edge devices.
  • Create optimization strategies for CPU, GPU, and NPU architectures.
  • Optimize for high performance and low power consumption in resource-constrained environments.
  • Analyze and benchmark performance across hardware and inference backends.
  • Use profiling tools such as NVIDIA Nsight Systems, Nsight Compute, and py-spy to identify GPU and runtime bottlenecks and improve inference performance.
  • Optimize trade-offs among model quality, inference performance, and cost for given prompts and service requirements, and design suitable serving architectures.

Desired Qualifications

  • Experience using or developing LLM serving frameworks such as vLLM, TensorRT-LLM, SGLang, llama.cpp, MLX, or ONNX Runtime.
  • Experience developing CUDA or Triton kernels or custom operators.
  • Experience with current LLM inference optimization and serving techniques such as speculative decoding, prefill-decode disaggregation, KV cache compression, or expert parallelism (MoE).
  • Experience optimizing NVIDIA GPUs, including H100 or B200, AMD GPUs, or other AI accelerators.
  • Experience optimizing edge AI and on-device LLMs for platforms such as Orin, Thor, Qualcomm, or Apple Silicon.
  • Experience with Kubernetes-based AI serving or operating large GPU clusters.
  • Understanding of LLM fine-tuning and distributed training.
  • Open-source contributions related to AI systems, ML systems, or LLM infrastructure.
  • Publications at leading AI/ML systems conferences such as NeurIPS, ICML, ICLR, MLSys, OSDI, NSDI, or ASPLOS, or equivalent research experience.

About the Company

42dot

Unified platform for multi-modal urban mobility

42dot.ai provides a platform called UMOS (Urban Mobility Operating System) that connects and manages multiple transportation modes—such as cars and drones—within a single system for cities, agencies, and private clients. UMOS enables efficient transportation management, demand-response services, and smart logistics by integrating data and operations across modes, powering operations through a subscription-based platform-as-a-service model. Unlike single-mode solutions, 42dot.ai offers multi-modal, city-focused orchestration that bridges public, private, and logistics needs. The company aims to transform urban mobility by helping cities run more efficiently, cut costs, and improve the experience for commuters and logistics providers.

Company size
501-1,000
Company stage
Growth Equity (Venture Capital)
Total funding
$731.9M
Headquarters
Seoul, South Korea
Founded
2019

Recently Posted Jobs

Explore more opportunities at 42dot

See more jobs