CareerPlan

Machine Learning Engineer - LLMs

AmsterdamFull-time2026-10-07 → 2026-11-19
Hybrid

Core

Develop domain-specific large language models (LLMs) by owning the full training pipeline from continued pre-training to advanced post-training techniques.

Role type

Senior Machine Learning Engineer (LLM Training & Infrastructure)

Builds

Production-grade language models and distributed training infrastructure for Prosus's AI-driven ecosystem.

Domain

Artificial Intelligence / Large Language Models / Distributed Systems

Deliverable

production ML models

Required skills

Large language model training (pre-training, fine-tuning), distributed training frameworks (DeepSpeed, FSDP, Megatron-LM, Axolotl), Python, PyTorch, data preparation (filtering, deduplication, mixture strategies), post-training techniques (RLHF, DPO, GRPO), evaluation frameworks, CI/CD

Preferred skills

Published research at ML conferences, synthetic data generation pipelines, model optimization for inference (quantization, distillation), memory optimization (gradient checkpointing, mixed precision), open-source contributions

Technologies

DeepSpeed, FSDP, Megatron-LM, Axolotl, PyTorch, Transformers, Accelerate, vLLM, H200 GPUs

Responsibilities

Design and execute experiments to improve model quality and training efficiency; Train LLMs using continued pre-training and full parameter fine-tuning; Build and optimize distributed training infrastructure across multi-node GPU clusters; Own large-scale data preparation and curate synthetic data; Debug training stability and optimize throughput; Implement techniques from recent research papers; Build robust evaluation frameworks.

Seniority

Senior, hands-on IC