Rubrics as Rewards: Reinforcement Learning Beyond Verifiable Domains | DocHero AI - Best paraphrasing and translation tool for academic and professional writing