Senior ML Engineer, Synthetic Data Pipelines (Remote)

Cohere; United States; Remote; USD 160,000 - 210,000; Full time

Benefits offered by this job

Weekly lunch stipend

Health and dental benefits

RRSP/401K matching

Parental Leave top-up

Education stipend

6 weeks paid vacation

Home office stipend

Job summary

Cohere is hiring a Senior Machine Learning Engineer focused on synthetic data to build and optimize the data pipeline for our language models. You will manage end-to-end synthetic data processes, including data generation, ablations, and quality evaluation, using diverse web and code data.

Remote-friendly with ET–CET overlap and global collaboration. You will bridge research and engineering to improve throughput and accelerator utilization, contributing to Cohere’s mission of efficient language

Qualifications

• Experience building data pipelines in Python.
• Familiarity with data processing frameworks such as Spark or Beam.
• Experience with LLMs through projects or open-source.
• Familiarity with LLM inference frameworks such as vLLM and TensorRT.
• Experience with large-scale multilingual corpora.
• Strong bridge between research and engineering to solve data challenges in AI training.

Responsibilities

• Design and maintain synthetic data pipelines for language models.
• Perform data ablations to assess data quality and experiment with data mixtures to enhance model performance.
• Research and implement innovative synthetic data curation methods.
• Collaborate with researchers and engineers to ensure data pipelines meet the demands of cutting-edge language models.

Skills

Python

Data pipelines

LLMs

Spark/Beam

TensorRT

Multilingual data

Research integration

Tools

vLLM

Pandas

Apache Spark

Apache Beam

Job description

Cohere is hiring a Senior Machine Learning Engineer focused on synthetic data to build and optimize the data pipeline for our language models. You will manage end-to-end synthetic data processes, including data generation, ablations, and quality evaluation, using diverse web and code data.

Remote-friendly with ET–CET overlap and global collaboration. You will bridge research and engineering to improve throughput and accelerator utilization, contributing to Cohere’s mission of efficient language