Rice must be carefully stored during the post-production stage, with an emphasis on controlling moisture content (MC) to preserve grain quality. This study presents an innovative method to identify MC and high moisture zones, along with irregularities in rice grain storage, employing cost-effective ultra-high frequency RFID (COTS UHF RFID) technology. Using a UHF RFID handheld reader operating at 915 to 919 MHz bands, experimental tests were conducted, to detect and locate high moisture zones in storage at 12, 16, 20, and 24% moisture levels. For density-independent MC detection and wet zone identification, variables such as RFID reader transmission power, operating frequency, storage materials, and MC location were methodically taken into consideration. The Received Signal Strength Indicator (RSSI) offers a powerful means to detect inconsistencies in moisture distribution within grain storage with the help of machine learning algorithms. Comparing the accuracy obtained from the validation test, RF and SVM are the algorithms that provide high accuracy greater than 70%. The small difference between precision and recall suggests that the distribution of data between each class is well-balanced. Based on the confusion matrix and learning curve, RF was found to be the best algorithm for the determination of rice MC and wet zone localization with an accuracy of approximately 74%, affirming the reliability of the proposed method. This study proves that the determination of MC using RFID technology and machine learning offers a portable and non-destructive method compared to state-of-the-art approaches that will damage the rice.