Design, implement, and evaluate RL training environments and reward signals; train and profile LLM post-training workflows; architect and scale distributed RL training infrastructure; optimize end-to-end experiment throughput and close the loop between environment design and model capability.