P

Prosus

Operates technology-led consumer internet ecosystems

ML Infrastructure & Acceleration Engineer - LLMs

Applied Machine LearningConfirmed live in the last 24 hours
Full-timeAmsterdam, NetherlandsNo salary listed

About the Job

Requirements

  • At least 5 years of experience in ML infrastructure, MLOps, HPC, or platform engineering focused on GPU workloads.
  • Hands-on experience operating multi-node GPU clusters with Kubernetes and/or Slurm, including debugging hardware, networking (NCCL, InfiniBand/RoCE), and job failures.
  • Production experience deploying and optimizing LLM inference with vLLM, SGLang, or TensorRT-LLM.
  • A proven track record of profiling and accelerating GPU workloads with measured results in throughput, latency, MFU, or cost.
  • A solid understanding of monitoring and observability for GPU systems, such as Prometheus, Grafana, or DCGM.

Responsibilities

  • Own the H200 GPU cluster end to end, using Kubernetes and Slurm for training and inference workloads with multi-tenant scheduling, quotas, and resource isolation.
  • Maintain the health of large distributed training runs through node health checks, NCCL and interconnect debugging, automated job recovery, and checkpoint management.
  • Track and optimize GPU utilization and cost across training and inference using dashboards and budget alerting.
  • Manage infrastructure-as-code and provisioning for GPU environments.
  • Define and monitor SLIs/SLOs for cluster availability and production model serving.
  • Deploy LLMs to production using vLLM or SGLang, with CI/CD, model versioning, and safe rollouts and rollbacks.
  • Monitor serving latency, throughput, errors, and cost, and act on regressions.
  • Profile training and inference end to end with PyTorch Profiler, Nsight Systems, and Nsight Compute to find and remove bottlenecks.
  • Improve training throughput and MFU using parallelism strategies (FSDP, tensor and pipeline parallelism), communication/compute overlap, mixed precision and FP8, activation checkpointing, and torch.compile.
  • Optimize inference using quantization (FP8, INT8, INT4), speculative decoding, prefix caching, continuous batching, KV cache tuning, and parallelism layouts.
  • Write or integrate fused and custom kernels with Triton or CUDA where off-the-shelf implementations fall short.
  • Build benchmarking tooling so every optimization ships with measured gains.
  • Partner with ML engineers and researchers to make training and serving efficient by default.
  • Establish best practices for using the cluster and guide teams in their adoption.

Desired Qualifications

  • Experience writing custom kernels in Triton or CUDA and familiarity with fused kernels such as FlashAttention.
  • Performance tuning of distributed training frameworks such as FSDP, Megatron-LM, or DeepSpeed.
  • Experience with low-precision training and inference (FP8, INT4) and model compression for production deployment.
  • Deep knowledge of GPU architectures, including memory hierarchy, tensor cores, and roofline analysis.
  • Experience with multi-tenant GPU scheduling and fair-share resource allocation.
  • Contributions to open-source projects such as vLLM, SGLang, PyTorch, or Triton.

About the Company

Prosus

Operates technology-led consumer internet ecosystems

Prosus is a global consumer internet group that builds technology-led lifestyle ecosystems across Europe, Latin America, and India, owning platforms like iFood, OLX, PayU, Swiggy, Meesho, and Just Eat Takeaway.com in 100+ markets. Its products are a network of apps and websites that connect consumers, merchants, and partners, powered by hands-on operations, investments, and AI-enabled technology to scale services. The company differentiates itself through its global footprint, diversified portfolio, and cross-market capabilities that support rapid growth across multiple sectors. Its goal is to expand its ecosystem to more markets, deepen engagement on its platforms, and create value by scaling trusted digital services for consumers, retailers, and partners.

Company size
10,001+
Company stage
IPO
Headquarters
Amsterdam, Netherlands
Founded
2015

Recently Posted Jobs

Explore more opportunities at Prosus

See more jobs