<![CDATA[Climate change, distribution disruptions, and global food price volatility have expanded the potential risks of food insecurity at the regional level. Provincial governments can no longer rely on conventional administrative approaches to respond to food system dynamics. Instead, they must develop information systems capable of processing data rapidly, adaptively, and accurately. This study explores how local governments in Indonesia—specifically Central Java and Bangka Belitung—formulate food crisis prediction strategies based on Big Data by integrating agricultural production data, rainfall levels, market prices, and public sentiment analysis. Using a qualitative-descriptive approach through document analysis, this research examines the architecture of data collection systems, the analytical models employed, and the institutional responses generated from data interpretation. The findings show that food insecurity predictions can be carried out more quickly and accurately when governments build a data ecosystem that integrates various information sources into a single, structured processing flow. Strategies such as regional food stock provision, prediction-based logistics distribution, and price interventions in vulnerable areas do not stem from manual reports but from systems that detect early signs of crisis. This study affirms that leveraging Big Data not only modernizes food governance but also shifts the government's orientation from reaction to anticipation, from administration to prediction, and from reporting to evidence-based intervention.]]>
