The work involves providing streamflow forecasts across short-term (10-day), seasonal (90-day), and annual (daily, 1-year) horizons to support risk-informed decision making. Forecasts must include points at Prado Dam inflow, historical reforecasts for validation in CSV format, extended quantiles for full hydrograph prescription, and probabilistic confidence intervals. The solution should combine AI-powered hydrological and meteorological forecasting with traditional models, delivering precise multi-horizon predictions from hourly to annual. The core purpose is to accurately predict rainfall events and basin inflow volumes that could trigger emergency protective measures in a construction project’s action plan. No timelines, budget amounts, or performance location were specified in the description.