Research Engineer / Scientist – Reinforcement Learning

πŸ•’ October 16

Where this job comes from

Posted on Prime Intellect's careers page β†—Prime Intellect's careers page, not through a job board

First seen by Remote Rocketship October 16

Still open on Prime Intellect's site. Checked 1 hour ago.

Apply Now
Logo of Prime Intellect

Prime Intellect

1 - 10 employees

πŸ€– Artificial Intelligence

☁️ SaaS

Artificial Intelligence β€’ SaaS β€’ Cloud Computing

Prime Intellect is a company focused on democratizing AI development by providing scalable and decentralized computing resources for training models. Their platform allows users to find and share global compute resources, enabling the training of state-of-the-art models through distributed clusters. They promote the collective ownership of AI innovations, including language and scientific models. Prime Intellect also offers a range of GPU options to facilitate affordable and efficient model training. They aim to advance decentralized training research and open-source AI development on a global scale.

πŸ“‹ Description

β€’ Develop and test RL methods for reasoning, coding, and long-horizon tasks, including changes to training objectives and policy optimization β€’ Investigate stability in asynchronous RL, including policy staleness, trainer/sampler mismatch, and numerical differences between training and rollouts β€’ Work with data and evaluation researchers to understand how task selection, rewards, and training data affect what the model learns β€’ Design experiments to understand how RL performance and learning dynamics change with model size, training compute, and task horizon β€’ Use controlled ablations and evaluations to distinguish learning gains from reward exploitation, evaluation artifacts, or training bugs β€’ Turn experimental findings into training recipes and test whether improvements hold across larger models and longer runs β€’ Extend prime-rl to support new algorithms, models, and asynchronous training workflows β€’ Improve performance and numerical correctness of research runs across trainers, rollout workers, and environments with the infrastructure team β€’ Build diagnostics that make unstable runs and performance regressions easier to reproduce and help distinguish algorithm problems from implementation or runtime issues

🎯 Requirements

β€’ Demonstrated depth in at least one of RL algorithms, empirical research on model training, or distributed ML systems β€’ Strong programming skills in Python and experience with PyTorch or a comparable ML framework β€’ Ability to design controlled experiments, interpret results, and debug unexpected model or system behavior β€’ Experience taking a research idea or systems change through implementation, testing, and evaluation, with a clear account of what worked and what didn’t β€’ Experience with distributed training frameworks or asynchronous rollout pipelines β€’ Experience diagnosing unstable RL runs, including policy-staleness or numerical issues β€’ Work on coding agents, verifiable rewards, or long-horizon environments

πŸ–οΈ Benefits

β€’ Equity incentives β€’ Flexible work arrangements, with the option to work remotely or in-person at our offices in San Francisco β€’ Visa sponsorship and relocation assistance for international candidates β€’ Quarterly team off-sites β€’ Hackathons β€’ Conferences and learning opportunities β€’ Opportunity to work with a talented, hard-working and mission-driven team

Apply Now

πŸ“Š Check your resume score for this job

Improve your chances of getting an interview by checking your resume score before you apply.

Similar Jobs

πŸ•’ October 9

Sequen

11 - 50

☁️ SaaS

πŸ€– Artificial Intelligence

🀝 B2B

Staff Research Engineer owning Sequen’s Python runtime for autonomous AI research agents. Building resilient infrastructure, evaluation systems, and observability for production ranking intelligence.

πŸ•’ October 9

Sequen

11 - 50

☁️ SaaS

πŸ€– Artificial Intelligence

🀝 B2B

Staff+ Research Engineer advancing Sequen’s autonomous ranking intelligence engine through recursive self-improvement. Setting research direction, running experiments, and shipping production improvements for major retailers and platforms.

πŸ•’ October 8

Burq

11 - 50

🍽️ Food & Beverage

πŸ₯ Healthcare

πŸ—οΈ Construction

Staff ML/OR engineer building ML-assisted dispatch, pricing, forecasting, and routing systems for Burq's logistics platform. Owning optimization architecture and production models with company-level impact.

πŸ•’ September 29

Higharc

11 - 50

πŸ—οΈ Construction

πŸ’Ό Consulting

🏠 Real Estate

Applied Research Engineer shipping spatial AI models into Higharc’s home-design and construction platform. Building APIs, evaluation pipelines, and production ML capabilities.

πŸ•’ September 25

829 Studios

201 - 500

πŸ“£ Marketing

🀝 B2B

πŸ€– Artificial Intelligence

AI Innovation Engineer building Python agentic workflows, MCP integrations, and AI tools for 829 Studios, a Boston-based digital marketing agency. Driving adoption across marketing operations.

πŸ‡ΊπŸ‡Έ United States – Remote

πŸ’΅ $110k - $135k / year

πŸ’° Series unknown on 2021-09

⏰ Full Time

🟑 Mid-level

🟠 Senior

πŸ“š Research Engineer

Developed byΒ Lior Neu-ner. I'd love to hear your feedback β€” Get in touch via DM or support@remoterocketship.com
Follow Remote Rocketship on LinkedInFollow Remote Rocketship on InstagramFollow Remote Rocketship on FacebookFollow Remote Rocketship on TikTok