RL / Post-Training Researcher

About the role

We're partnering with a well-funded AI Fintech Lab tackling some of the most complex real-world problems at the intersection of machine learning, reasoning, and large-scale systems. The team is looking for a Research Scientist focused on reinforcement learning and post-training to help advance model capabilities beyond foundation-model baselines.

You'll work on problems spanning RLHF/RLAIF, preference optimization, reward modeling, synthetic data generation, evaluation frameworks, reasoning enhancement, and large-scale post-training pipelines. The work sits at the boundary of research and production, with a direct path from experimentation to deployment.

This is an opportunity to collaborate with a small group of high-caliber researchers and engineers building novel training systems, developing new approaches to model alignment and optimization, and pushing the limits of what modern agentic AI systems can achieve in real-world environments.

Ideal Experience

• Frontier AI research and engineering
• Reinforcement learning, RLHF, RLAIF, DPO, PPO, GRPO, or reward modeling
• LLM post-training and model alignment
• Evaluation frameworks and model benchmarking
• Synthetic data generation and data flywheel development
• Deep learning and large-scale distributed training

The team is particularly interested in candidates from frontier AI labs, top-tier research organizations, leading technology companies, or high-performing AI startups.

Qualifications

• Degree in CS or a related field from a top-tier institution

Pay

Compensation will be highly competitive to frontier labs, including cash base, discretionary bonus, and equity.