In the 70s, the interplay between microscale electronics and mechanics gave birth to micro- and nanoelectromechanical systems (MEMS/NEMS) that are prevalent in our daily life. The emergence of silicon photonics in the 90s was a result of the marriage between microelectronics and optics promising extreme communication bandwidth and processing power. A few years ago, the field of microscale optomechanics that harnesses the interaction between light and mechanics on a nanoscale emerged. The field witnessed the birth of many exciting technologies as quantum limited detection of ultra-weak forces, preparation of micromechanical oscillators close to their motional quantum ground states and enabling self-sustaining oscillations of mechanics with light. The aim of this thesis is to explore and address a few challenges in coupled optomechanical systems. So far, most work in this area focuses on single device behaviors. One could imagine that like connecting many transistors together leads to complex computing machines, a network of coupled optomechanical devices have the potential to offer dynamics that are not accessible with single optomechanical devices. In this thesis, I show that indeed, light can be used to synchronize arrays of mechanical oscillators even when they are not physically connected. I will also show in this thesis that coupling distinct optical and mechanical elements together could also enable a new paradigm of devices. We couple a single Carbon Nanotube (CNT) strongly to on-chip high-Q optical microcavi- ties. Despite the tiny size of CNT, we show that the optical microcavity is still extremely sensitive to the CNT motion. We demonstrate that we can observe in real-time the thermal Brownian motion of a single CNT for the first time. The unique carbon-optical system also enables an almost completely dissipative optomechanical system that has not been achieved in any other type of systems to date.