We aim to decrease the interactivity boundary between humans and robots through human motion mimicking to create meaningful motions for social robots. This paper presents a realtime human motion mimicking system that utilizes equipment commonly available and easily accessible to program robots to mimic human motion. Using a combination of lightweight artificial intelligence models, low to medium end computing power, and low cost peripherals, we can create a system that performs motion mimicking in realtime. The performance of this system is observed using the SoftBank Robotic’s Pepper humanoid robot. The current system is sufficient as a simple motion mimicking system and demonstrates the overall system pipeline. However, with the current trajectory of AI and machine learning, further advancements in the tools utilized can significantly improve the proposed system and expand its potential use cases.