Research Engineer – Reinforcement Learning
About the Role
A high-impact AI research initiative focused on advancing reinforcement learning capabilities through the development of specialized environments, scalable training pipelines, and automated evaluation systems. The work combines applied research with production-grade engineering to improve the performance, reliability, and capabilities of modern AI models.
This opportunity is ideal for experienced Research Engineers with strong expertise in reinforcement learning, ML-oriented data design, environment development, and model training workflows. Candidates should be comfortable translating research concepts into practical systems and building reproducible experimentation frameworks.
The work involves designing RL environments, developing synthetic data generation pipelines, fine-tuning open-source models, and implementing automated evaluation systems. Strong engineering execution, experimental rigor, and the ability to measure model performance reliably are critical.
What You'll Do
- Architect self-contained reinforcement learning environments for complex real-world tasks.
- Develop reward functions, verification mechanisms, and evaluation logic.
- Design and scale episode-generation pipelines and multi-component training workflows.
- Build automated synthetic data generation systems to support model training.
- Develop AI-driven evaluation, grading, and quality assurance systems.
- Fine-tune and optimize open-source reinforcement learning models using custom datasets.
- Implement benchmarking frameworks to measure model capabilities, robustness, and data quality.
- Analyze model performance across different evaluation tasks and benchmarks.
- Establish reproducible experimentation and model validation workflows.
- Contribute to the development and technical analysis of internal and external evaluation benchmarks.
Requirements
- Deep understanding of reinforcement learning principles, algorithms, and training dynamics.
- Practical experience designing and developing RL environments.
- Strong knowledge of ML-oriented data design and reinforcement learning workflows.
- Experience building and scaling RL training pipelines or experimentation frameworks.
- Proficiency in automation and synthetic data generation.
- Familiarity with automated model evaluation, validation, and quality assurance systems.
- Experience fine-tuning and evaluating open-source machine learning models.
- Ability to design reliable reward functions, verifiers, and benchmarking systems.
- Strong analytical thinking and problem-solving skills.
- Clear technical communication and documentation abilities.
- Ability to work independently in fast-paced, research-driven environments.
- Experience publishing benchmarks, evaluations, or research artifacts is preferred.
- Familiarity with established model evaluation ecosystems and benchmark frameworks is preferred.
- Background in scalable infrastructure for large-scale reinforcement learning experimentation is preferred.