jobapplier

Machine Learning Engineer Intern

Coinbase

Skills

Required

  • machine learning
  • ML
  • software engineering
  • version control
  • Python

Preferred

  • learning / reinforcement learning
  • deep learning
  • structured or unstructured datasets
  • machine learning models and pipelines
  • building and training models
  • PyTorch / TensorFlow

Coinbase's Machine Learning team as a Machine Learning Engineer Intern for a 12-week summer internship in 2027. You'll work alongside senior engineers to build ML models and pipelines that enhance platform security, personalize user experiences, and unlock new crypto use cases.

Take ownership of a research-to-production project, applying cutting-edge ML techniques to real-world problems at blockchain scale.

About Coinbase

Partner with senior engineers and product teams to identify new ML applications for blockchain and crypto use cases, and present findings to stakeholders.

Responsibilities

  • You'll take ownership of a research-to-production project, develop and deploy machine learning models, and present your findings to cross-functional stakeholders.
  • Develop, deploy, and operate machine learning models and pipelines at production scale, driving an end-to-end research project.
  • Demonstrated proficiency building and training models using ML frameworks such as PyTorch or TensorFlow

Requirements

  • Candidates should be pursuing a Ph.D. in machine learning or a related field and have experience with ML frameworks and software engineering fundamentals.
  • Utilizes generative AI responsibly, maintaining human oversight to deliver business-ready outputs and drive measurable improvements in workflow efficiency, cost, and quality
  • Familiarity with software engineering fundamentals including version control, testing, and writing production-quality Python code

Preferred

  • Currently pursuing a Ph.D. with published or in-progress research in machine learning, deep learning, or a closely related field
  • Experience applying ML techniques including supervised learning, unsupervised learning, or reinforcement learning to structured or unstructured datasets