In the past years, the advancements in computer graphics have brought an abundance of applications with superb virtual characters that are closing the visual gap of the uncanny valley. Machine learning has also experienced a breakthrough with new deep learning techniques. With today’s computational power allowing graphics and artificial intelligence in real-time, virtual intelligent agents combining these two features are already on the front doorstep. Although intelligence can be perceived in many distinct ways, here we address the behavioral animation of these intelligent agents. Moreover, we approach this problem using a hierarchical solution benefiting from traditional artificial intelligence and newer deep learning. In turn, we bring to life agents intended to get closer to Artificial Life instead of ’wooden’ automation agents that do not take account of human perception.