Rapid post-event flood mapping can support infrastructure response and federal reimbursement processes that currently take two to five days. Researchers are developing and operationally validating a decision-support system for El Paso Water that uses fine-tuned Prithvi-EO foundation models and Harmonized Landsat-Sentinel-2 imagery to produce flood maps within 24 to 48 hours. Work will advance the system from proof of concept to demonstration in an operational environment through benchmarking against Convolutional Neural Network models, staged validation, and testing during monsoon events. Deliverables include curated training datasets, open-source code, uncertainty quantification protocols, and transition and sustainability plans. Results may reduce assessment delays and improve information available to El Paso Water and regional infrastructure and emergency-response organizations.