Humans, like many other species, heavily rely on social interactions for survival. Yet our understanding of the brain mechanisms underlying the development of social behavior is strikingly limited. This knowledge gap poses significant challenges for developing treatments for disorders marked by social impairments, such as the majority of neuropsychiatric and neurodevelopmental conditions. The significance of our social environment makes it even more detrimental if the ability to fully engage with it is impaired. Current methods for analyzing social behavior in neuroscience are often artificial and constrained, failing to consider crucial elements like the social milieu and intrinsic factors. To address this challenge, we developed an unbiased multimodal strategy leveraging machine learning and genetic mouse models of ASD. By observing rodents in their natural environment over the long term, we aim to capture essential determinants of social skills. Our project seeks to elucidate how social context and internal variables influence the development of rodent social abilities and the underlying neural circuits. Understanding these dynamics could lead to novel treatment targets for conditions characterized by social deficits by including the environment as a contributing factor.