In the Philippines, about twenty (20) typhoons occur annually, causing heavy flooding which poses risks that lead to injuries and casualties despite preparedness measures. This study addresses the problem of hindered rescue efforts due to limited resources, dangerous access to flooded areas, and damaged communication infrastructures by introducing an innovative solution: an unmanned amphibious robot for search and monitoring tasks. The developed robot is capable of locating human presence and help needed while providing a live video feed. Evaluations demonstrated the capabilities of the robot to navigate both on land and water with respective speeds of 1.2 m/s and 0.205 m/s over a 120-m LoRa communication. The live video feed quality highlights the feasibility of a 4G LTE network for real-time display. The trained YOLOv5 model had high accuracy in detecting human presence and help needed over 3.5m and 7m distances with 90% and 93.33%, respectively. GPS coordinate reception yields good results in open areas only. There was also a seamless integration of data from the robot to the local website, offering accessible data. Limitations arose when live video feed streaming and YOLOv5 processing were performed simultaneously. This research contributes to aiding post-typhoon heavy flooding response by developing an unmanned amphibious robot, offering insights into its performance and potential for real-world applications in disaster response scenarios.