This study tackles waste management challenges in Costa Rica, where urban-rural disparities limit access to recycling facilities, posing barriers to national sustainability goals. By analyzing optimal facility placements, this research aims to improve accessibility and streamline waste management using Agent-Based Modeling (ABM) and Reinforcement Learning (RL). Integrating geospatial and demographic data, it forecasts recycling behaviors across Costa Rican provinces, with RL identifying cost-effective facility locations to boost recycling rates. Findings reveal notable accessibility gaps—urban centers like San José achieve a 56% accessibility rate, whereas rural areas, such as Puntarenas, fall below 30%. Strategic placements could elevate Cartago’s recycling rate to 27.38%, and RL optimization indicates a 53.4% potential rise in recycling and a 12% decrease in landfill reliance in accessible areas. These outcomes underscore the need for region-specific investments in waste infrastructure, suggesting that AI-enhanced waste categorization and community engagement could further progress Costa Rica’s sustainable waste management.