An automatic feeding system that can be monitored and managed remotely is needed amidst the busy schedules of pet owners, especially for cats. As a result, an automatic cat feeder that can be operated and viewed remotely via the Internet of Things was developed in this research. By understanding the properties of loadcell sensors and servo motors, research in this case seeks to create an automatic cat feeding system that uses a PID approach to regulate food according to the cat's needs. For the convenience of remote use, the system also allows monitoring and control using the Telegram application. An ESP32 microcontroller, a servo for food release, a loadcell sensor for food weight measurement, and a camera for cat detection were all used in the development of this research system. An automatic cat feeding system that produces servo motor and load cell sensor calibration values with average error results of 1.84%, 3.86%, and accuracy of 96.14%, 98.16% was successfully developed in this research. With an error percentage of 1.67%, this research was able to produce food that matched the cat food dosage using the PID approach with trial and error tuning, especially with a cat food dosage of 60 grams. Based on the research findings, it can be said that this system works as planned and offers a practical way to feed cats food precisely and automatically.