San Francisco, CA

Research Scientist, Discovery (Engineering & Applied Sciences)

USD 500K–850K / yearHybridData and machine learning

PythonPyTorchLLM

Original posting ↗Apply ↗Company job board ↗Follow new Python jobs in Telegram

Anthropic, a software company in AI and data, has an open position for Research Scientist, Discovery (Engineering & Applied Sciences). It is a hybrid position in San Francisco, United States. We classify it under data and machine learning. The stack in the posting includes PyTorch, LLM and Python. Visa sponsorship is offered. There are 169 open positions from Anthropic on this site, 52 of them in IT and engineering.

Posted today · Last checked 2026-11-05

About Anthropic

AI and data Software product company 169 open positions, 31% in IT and engineering. Most open roles: Software engineering (general) (28), Sales (22), General management and executives (19).

All open jobs at Anthropic →

Similar jobs

Sr Software Engineer - Machine Learning & Data ClassPass · United StatesUSD 150K–200K / year Staff Machine Learning Engineer, AI InfrastructureWaymo · Mountain View, CA, USA; San Francisco, CA, USA; New York, NY, USAUSD 251K–310K / year Lead Bioinformatics & Machine Learning Scientist (Epigenomics)Natera · San Carlos, CA Data Engineer II, JVMPlayStation Global · United States, San Mateo, CAUSD 163.7K–245.5K / year Senior Consultant Data AnalystExcella · Arlington, Virginia, United States, RemoteUSD 98K–144K / year Machine Learning EngineerReddit · San Francisco, CAUSD 216.7K–303.4K / year Senior Data ScientistReddit · San Francisco, CAUSD 229K–267.1K / year Senior AI Engineer (CCC Auto / Agentic)EvolutionIQ · Chicago, IL or RemoteUSD 180K–225K / year

1,278 open data and machine learning jobs in the United States All open jobs at Anthropic →

Workplace	Hybrid
Seniority	Not stated
Role	Data and machine learning
Countries	United States
Languages	Python
Key facts	Stack: Python, PyTorch, LLM · Visa sponsorship: yes

The full description is on the original posting. Labels are set automatically and can be wrong. About the data