Skip to content
jobs·jobs·jobs.com

Remote Machine Learning Scientist jobs

9 open positions in the United States, newest first.

Develops new machine learning models and methods, working closer to research than a typical data scientist or ML engineer. The machine learning scientist reads and extends the literature, designs experiments, trains and evaluates models, and works with engineers to bring the successful ones into products. Problems are often open-ended, such as improving a recommender, a language model, or a forecasting system beyond known baselines.

About the Machine Learning Scientist role

What the work looks like

  • Formulates a modeling problem, selects baselines, and defines offline metrics that reflect the real objective.
  • Implements and trains models in PyTorch or JAX on GPU clusters and runs ablation studies.
  • Reads recent papers, reproduces promising results, and adapts them to the company's data and constraints.
  • Analyzes model errors and data quality issues and designs the next round of experiments.
  • Writes up findings internally or for publication and hands working models to engineers for productionization.

What employers ask for

  • PhD or equivalent research experience in ML
  • PyTorch, JAX, or TensorFlow
  • Strong grounding in statistics and optimization
  • Experiment design and rigorous evaluation
  • Python and distributed training on GPUs
  • Publication record, often valued

Doing this job remotely

Experiments run on cloud or cluster compute, so the work itself is location-independent. Research benefits from frequent discussion, and remote scientists rely on written experiment logs, reading groups, and regular reviews to stay in step. Access to proprietary data and models is tightly controlled, and some research labs prefer staff near a hub office.

Where it comes from and where it leads

Most hold a doctorate or master's in machine learning, statistics, physics, or a related field, or moved over from ML engineering with research results. They advance to senior and principal scientist, research lead, or applied science management.