Remote AI & Machine Learning Jobs
216 Remote Jobs available
Filters
Showing 141–160 of 216
- Last monthView job
- GitLabLast month4 countriesUSD 250K–325K / yearPrincipal#PrototypingLast monthView job
- SamsaraLast monthUnited StatesUSD 170K–286K / yearSenior#AWS#Azure#C++Last monthView job
- MercuryLast monthCA · USCAD 167K–197K / yearSenior#Airflow#CI/CD#dbtLast monthView job
- Last monthView job
- Signifyd2 months ago2 months agoView job4 Day Week
- Oak Ridge National Laboratory2 months ago2 months agoView job4 Day Week
- 2 months agoView jobRemote OK
- 2 months agoView job
- 2 months agoView job
- Instacart2 months agoCanadaCAD 180K–190K / yearSenior#pandas#Python#scikit-learn2 months agoView job
- Instacart2 months agoUnited StatesSenior#pandas#Python#scikit-learn2 months agoView job
- 2 months agoView job
- Descript2 months agoCA · USUSD 197K–263K / year#Computer Vision#LLMs#PyTorch2 months agoView job
- Discord2 months agoUnited StatesUSD 272K–340K / yearManager2 months agoView job
- 2 months agoView job4 Day Week
- Affirm2 months agoCanadaManager#Python#SQL2 months agoView job
- Affirm2 months agoUnited StatesUSD 227K–287K / yearManager#Python#SQL2 months agoView job
- 2 months agoView job
- 2 months agoView job
Machine learning roles cover training, evaluation and deployment of models, and increasingly the applied work of building products on top of foundation models: retrieval, evaluation harnesses, prompt and agent design, and inference cost.
AI labs and the wave of companies building on their APIs, plus established firms adding ML to search, fraud and forecasting. The applied end of this market has grown far faster than the research end.
Separate research from applied engineering before you apply — the two share a title and share very little else. Check whether the team owns models in production or hands them to another group, and what the evaluation story is.
Common questions
- Do these roles need a PhD?
- Research positions frequently do. Applied and infrastructure roles building on existing models usually do not, and increasingly say so. The posting is explicit far more often than not — read the requirements rather than assuming.

