← All roles

Applied Research - Agentic RL

Prime Intellect· Applied Research
Apply on company site
San Francisco, New York, Remote, London$150k - $300kFull-timePosted 4d ago

About the role

Role focused on reinforcement learning for agents that reason over long horizons, use tools, coordinate with other agents, recover from failure, and improve from experience. Cross-domain applied research in RL and ML systems with emphasis on algorithms, environments, harnesses, reward/credit-assignment, evaluation, and scalable experimental systems. Externally facing role collaborating with application teams, partners, and open-source users.

Responsibilities

  • Applied Research at Prime: turn workflows into tasks, environments, evaluations, and reward signals; run post-training experiments against real baselines; dissect gains to model, data, harness, inference, or infrastructure; reuse parts as shared research, open-source systems, and product capabilities.
  • Role scope: extend Prime's verifiers, algorithms layer, multi-agent training, Prime Agent, and RL systems tested at trillion-parameter scale; pressure-test against live application programs.
  • Representative early problems: - Build agentic judges and credit-assignment methods inspecting complete trajectories, using tools, and providing learning signals beyond a single final score. - Make multi-agent and long-horizon environments train reliably at scale with debuggable traces and baselines. - Detect when gains come from reward hacking, harness behavior, data leakage, or infrastructure artifacts; redesign loop for robust results.
  • What you will own: - RL and post-training methods for tool-using, multi-agent, memory-augmented, and long-horizon systems. - Reusable environments, harnesses, rewards, and evaluations. - Experiments linking algorithm choices to training/inference behavior. - Open-source systems and papers derived from problems surfaced by application researchers and external users.

Qualifications

  • Relevant experience: hands-on RL training, scalable ML systems, agent environments, evaluation, inference/training infrastructure, or research crossing methods and systems.
  • Hands-on experience building and debugging model-training, post-training, agent, evaluation, or closely adjacent experimental loops.
  • Strong experimental judgment: meaningful baselines, careful trace inspection, clear failure attribution, and willingness to pivot as evidence evolves.
  • Strong Python and ability to build artifacts that other researchers and engineers can run, inspect, and extend.
  • Comfort owning ambiguous problems and keeping partners/teammates updated as you resolve them.

Skills

Required:Reinforcement LearningPost-training systemsPythonExperiment design and analysis
Nice to have:Systems thinking

Benefits

  • Cash Compensation Range of $150-300k + equity incentives
  • Flexible Work (San Francisco, New York, London, or remote)
  • Visa Sponsorship & relocation support
  • Professional Development budget
  • Team Off-sites & conference attendance
Last updated Oct 21, 2026. Interested? Apply on the company site.

About Prime Intellect

Commoditizing Compute & Intelligence

Prime Intellect democratizes AI development at scale. Our platform makes it easy to find global compute resources and train state-of-the-art models through distributed training across clusters. Collectively own the resulting open AI innovations, from language models to scientific breakthroughs.

AI developmentbessemer-ai-infrastructure-2026compute resourcesdistributed trainingmodelsreinforcement-learning-platformsscience
Company profileWebsite

More open roles at Prime Intellect