On this paper, the design, control and implementation of a robotic system for bartender services is realized. In this approachment, initially the QFD design methodology is used, so the product's final users requirements identification and prioritization could be done, alongside with the CAD and the fabrication of a functional 5 DOF prototype robot that performs manipulation tasks for drinks in collaboration with a camera on its final effector that supplies live vision of the surroundings. The proposed goal is making use of artificial vision and inverse kinematics in order to position the robot so that it could do some tasks meant to waiters. The 2D artificial vision system used was the fast detector and descriptor SURF (Speeded Up Robust Features) and servo-piloting for the camera to track and find interest objects and grasp them.