AI Engineer – LLM Data

About the role

The Institute of Foundation Models is a research lab focused on building, understanding, using and risk-managing foundation models. As an AI Engineer specializing in LLM data, you will build and improve high-quality training data for foundation models across pre-training, mid-training and post-training. The work includes large-scale data curation and processing, LLM-based data synthesis, data quality evaluation and experimentation on how data choices affect model performance.

Key responsibilities

• Build, curate and improve large-scale datasets for LLM pre-training, mid-training and post-training
• Support time-sensitive data and model development tasks in a fast-moving research environment
• Develop data processing pipelines covering extraction, cleaning, filtering, deduplication, quality scoring, transformation and dataset composition
• Design LLM-based data synthesis and augmentation pipelines, including prompt-based generation, filtering, refinement and quality control of synthetic data
• Research methods to improve training data quality, diversity, coverage and efficiency
• Run experiments linking training data to model performance and use evaluation results to guide data improvements
• Develop scalable tools and workflows for processing and analysing large datasets
• Analyse datasets statistically and with model-based methods to find quality issues, biases, duplication and distributional gaps
• Collaborate with researchers, model engineers and data teams, and document datasets, experiments and methodologies

Required profile

• Strong programming skills in Python and experience building engineering or research workflows
• Hands-on experience with machine learning, deep learning, NLP or large language models
• Experience with large-scale datasets, including processing, analysis, filtering, transformation and quality control
• Ability to investigate data or model quality issues independently and make data-driven technical decisions
• Strong problem-solving, communication and collaboration skills in a fast-paced AI research environment

Required skills

• Familiarity with ML and LLM tools such as PyTorch, Hugging Face, vLLM or similar
• Preferred: experience preparing data for LLM pre-training, fine-tuning, preference optimisation or RL; synthetic data generation with LLMs; LLM evaluation and benchmarking; distributed data processing and compute infrastructure