Job Description

Employment Type: Full-Time Location Type: Remote Salary: (Not specified)

Who are we?

Our mission is to scale intelligence to serve humanity. We’re training and deploying frontier models for developers and enterprises who are building AI systems to power magical experiences like content generation, semantic search, RAG, and agents. We believe that our work is instrumental to the widespread adoption of AI.

We obsess over what we build. Each one of us is responsible for contributing to increasing the capabilities of our models and the value they drive for our customers. We like to work hard and move fast to do what’s best for our customers.

Cohere is a team of researchers, engineers, designers, and more, who are passionate about their craft. Each person is one of the best in the world at what they do. We believe that a diverse range of perspectives is a requirement for building great products.

Join us on our mission and shape the future!

Why this role?

As a Pre-Training Data Engineer, you will play a pivotal role in developing the data infrastructure that underpins Cohere’s advanced language models. Your responsibilities will encompass the end-to-end management of training data, including ingestion, cleaning, filtering, and optimization, as well as data modeling to ensure datasets are structured and formatted for optimal model performance. You will work with diverse data sources—such as web data, code data, multilingual corpora, and synthetic data—to ensure their quality, diversity, and reliability.

In this role, you will design and implement scalable, robust pipelines for data processing, conduct data ablations to evaluate quality, and experiment with data mixtures to enhance model performance. By combining research and engineering, you will bridge the gap between raw data and cutting-edge AI models, directly contributing to improvements in critical training metrics like throughput and accelerator utilization.

Your work will be essential to Cohere’s mission of delivering efficient and reliable language understanding and generation capabilities, driving innovation in natural language processing. If you are passionate about transforming data into the foundation of AI systems, this role offers a unique opportunity to make a meaningful impact.

Please Note: We have offices in London, Paris, Toronto, Ottawa, San Francisco and New York but also embrace being remote-friendly! There are no restrictions on where you can be located for this role.

As a Data Engineer in the Pre-Training team, you will:

• Design and build scalable data pipelines to ingest, clean, filter, and optimize diverse datasets, including web data, code data, multilingual corpora, and synthetic data.
• Conduct data ablations to assess data quality and experiment with data mixtures to enhance model performance.
• Develop robust data modeling techniques to ensure datasets are structured and formatted for optimal training efficiency.
• Research and implement innovative data curation methods, leveraging Cohere’s infrastructure to drive advancements in natural language processing.
• Collaborate with cross-functional teams, including researchers and engineers, to ensure data pipelines meet the demands of cutting-edge language models.

You may be a good fit if you have:

• Strong software engineering skills, with proficiency in Python and experience building data pipelines.
• Familiarity with data processing frameworks such as Apache Spark, Apache Beam, Pandas, or similar tools.
• Experience working with large-scale datasets, including web data, code data, and multilingual corpora.
• Knowledge of data quality assessment techniques and experimentation with data mixtures.
• A passion for bridging research and engineering to solve complex data-related challenges in AI model training.

Bonus: Paper at top-tier venues (such as NeurIPS, ICML, ICLR, AIStats, MLSys, JMLR, AAAI, Nature, COLING, ACL, EMNLP).

If some of the above doesn’t line up perfectly with your experience, we still encourage you to apply! If you want to work really hard on a...