Applied Researcher: Agentic RL & Multi-Agent Systems

Prime Intellect AI; San Francisco (CA); Remote; USD 150,000 - 300,000; Full time

Benefits offered by this job

Flexible work locations (SF, NY,London

Visa sponsorship & relocation

Development budget

Team off-sites & conferences

Job summary

Prime Intellect AI is building the open frontier AI stack and seeking researchers to advance reinforcement learning for agents that reason over long horizons, use tools, and coordinate with other agents. This Applied Research role focuses on RL and ML systems across algorithms, environments, and scalable experiments.

You will work with application teams and open-source users to extend methods and deploy reusable research components that transfer across domains, with strong emphasis on rigorous

Qualifications

• Hands-on RL training and scalable ML systems experience.
• Experience building and debugging model-training or evaluation loops.
• Strong experimental judgment and ability to attribute failures.
• Strong Python and ability to build runnable artifacts.
• Comfort with ambiguous problems and communicating with partners.

Responsibilities

• Develop RL and post-training methods for tool-using, multi-agent, memory-augmented, and long-horizon systems.
• Create reusable environments, harnesses, rewards, and evaluation protocols.
• Run experiments linking algorithm choices to training and inference behavior.
• Publish open-source systems and papers from problems surfaced by application researchers and external users.
• Collaborate with application researchers and external users.

Skills

Reinforcement learning

ML systems

Agent environments

Python programming

Tools

Python

Job description

Prime Intellect AI is building the open frontier AI stack and seeking researchers to advance reinforcement learning for agents that reason over long horizons, use tools, and coordinate with other agents. This Applied Research role focuses on RL and ML systems across algorithms, environments, and scalable experiments.

You will work with application teams and open-source users to extend methods and deploy reusable research components that transfer across domains, with strong emphasis on rigorous