Cohere is hiring a Member of Technical Staff Pre Training Data in Toronto!

Cohere, an enterprise-focused AI company building frontier foundation models, is hiring a Member of Technical Staff for its pretraining data team. The role sits at the intersection of research and engineering: you would own and improve the data pipeline feeding Cohere's large language models, running experiments to judge data quality and assembling pretraining mixtures that lift model performance. Day-to-day work spans data ablation studies, data modeling and formatting for efficient training, and inventing new curation approaches that take advantage of Cohere's internal infrastructure. Success shows up in concrete training metrics such as throughput and accelerator utilization. The position is remote-friendly with offices in Toronto, London, Paris, New York, San Francisco and elsewhere; candidates can be based anywhere between Eastern Time and EU time zones, and no minimum in-office commitment is required. It suits someone who enjoys moving fluidly between experimentation and production data engineering, and who wants direct influence on how frontier models are trained.

Key responsibilities

  • Design, build and maintain the pretraining data pipeline used for Cohere's language models
  • Run data ablation studies to measure dataset quality and test alternative data mixtures
  • Apply data modeling techniques so datasets are structured and formatted for training efficiency
  • Research and implement new data curation methods using Cohere's infrastructure
  • Partner with researchers and engineers across teams to keep pipelines aligned with model training needs
  • Track and improve training metrics such as throughput and accelerator utilization

Key requirements

  • Strong software engineering background with proficiency in Python and hands-on data pipeline development
  • Familiarity with curriculum learning, data mixing and data attribution concepts
  • Experience with data processing frameworks such as Apache Spark, Apache Beam, Pandas or similar
  • Track record working with large-scale datasets, including web, code and multilingual corpora
  • Knowledge of data quality assessment and experimentation with data mixtures
  • Motivation to bridge research and engineering on complex AI training data problems
  • Bonus: published research at top-tier venues such as NeurIPS, ICML, ICLR, ACL, EMNLP or MLSys
  • Seniority: Mid-to-senior individual contributor (Member of Technical Staff)
  • Work mode: Remote-friendly; office or remote, with core hours between EST and EU time zones
  • Key skills: Python, Data pipelines, Apache Spark, Apache Beam, Pandas, Data curation, Data ablation, Curriculum learning, Data mixing, Data attribution, Large-scale datasets, Multilingual corpora





Want more jobs?


Join our mailing list to receive updates about new data science and Member of Technical Staff Pre Training Data jobs.



Twitter logo   Facebook logo   Instagram logo   Linkedin logo   Mastodon logo


LATEST DATA SCIENCE JOBS
    -    MAILING LIST
    -    POST A JOB
    -    RESOURCES