PARL (Parallel-Agent Reinforcement Learning) - A training paradigm for coordinating multiple agents in parallel workflows