Door lock security system based on voice recognition with natural language processing (NLP) is an innovation and innovative solution in improving home security and comfort using the internet of things (IoT). This research aims to develop and implement a security system that can recognize and process sentence variations to open or close doors with high success rate. The system uses ESP32 Dev Kit V1 microcontroller, MQTT communication protocol to send data, and integrates relays, solenoid door locks so that the system can operate efficiently and reliably. System testing shows that voice recognition and NLP can identify voice commands in human a special card or tag. The test results show that the system can accurately control the solenoid door locks based on the received commands. This research can contribute to the development of a more advanced and convenient home security systems.