AI Engineer – LLM Data

Build, curate, and improve large-scale datasets for LLM pre-training, mid-training, and post-training.

Rapidly support time-sensitive data and model development tasks in a fast-moving research environment.

Develop and improve data processing pipelines including data extraction, cleaning, filtering, deduplication, quality scoring, transformation, and dataset composition.

Design and implement LLM-based data synthesis and augmentation pipelines, including prompt-based generation, filtering, refinement, and quality control of synthetic data.

Research and apply methods for improving training data quality, diversity, coverage, and efficiency.

Design experiments to understand the relationship between training data and model performance, and use model evaluation results to guide data improvements.

Develop scalable tools and workflows for processing and analyzing large datasets efficiently.

Analyze datasets using both statistical and model-based methods to identify quality issues, biases, duplication, distributional gaps, and opportunities for improvement.

Collaborate closely with researchers, model engineers, and other data teams to translate model development needs into effective data solutions.

Document datasets, experiments, data processing methodologies, and key findings clearly to support reproducibility and knowledge sharing.

Experience preparing data for large-scale LLM pre-training, continued/mid-training, supervised fine-tuning, preference optimization, reinforcement learning, or other post-training workflows.

Experience with synthetic data generation using LLMs, including generation, filtering, verification, or quality evaluation.

Experience designing or running LLM evaluations, benchmarks, model training, or fine-tuning experiments.

Understanding of how data quality, mixture, diversity, and scaling affect foundation model performance.

Experience with large-scale or distributed data processing and compute infrastructure.

Experience working with research teams on rapidly evolving foundation model or generative AI projects.

Contributions to open-source AI/ML projects, relevant publications, or demonstrated hands-on work with modern foundation models are a plus.