Research Scientist

A scientist who codes – advancing RL with new, generalizable ideas.

### About the Role

As a Research Scientist at Void Main Lab, you will explore and develop reinforcement learning (RL) methods, synthetic data generation, and agent architectures that enable AI systems to not only imitate but invent. You will collaborate with a cross-functional team of researchers and engineers to push the boundaries of AI-native creation in video, gaming, and software.

### Responsibilities

- Conduct original research in reinforcement learning, reward design, and agent-based systems.

- Design and implement algorithms that generalize across domains and tasks.

- Explore new paradigms for synthetic data generation and long-context learning.

- Work closely with product and engineering teams to translate research into practical applications.

- Publish high-quality research in top conferences and journals, and contribute to the global research community.

- Drive innovation by staying ahead of emerging trends in AI, ML, and computational creativity.

### Requirements

- PhD in Computer Science, Machine Learning, Mathematics, or related field, or a Master’s degree with outstanding research/industry experience .

- Strong background in reinforcement learning, optimization, or probabilistic modeling .

- Proficiency in Python and at least one major ML framework (e.g., PyTorch, JAX, TensorFlow).

- Strong publication record in top AI/ML venues (e.g., NeurIPS, ICML, ICLR, AAAI, CVPR).

- Ability to design, implement, and evaluate new algorithms from scratch.

- Excellent problem-solving skills, creativity, and a drive for pushing the boundaries of AI.

- Strong communication and collaboration skills; ability to work in a fast-paced, research-driven environment.

Preferred:

- Experience with large-scale distributed training and experimentation.

- Familiarity with generative models, multimodal learning, or self-play environments.

- Passion for exploring the frontier of AI creativity and building AI-native systems.