This paper introduces a methodology for predicting optimal temporal intervals during the day and the day-of-week that are conducive to enhancing the visibility and interaction of social media video posts, within the domain of Instagram, through the careful integration of temporal analytics and machine learning paradigms. The conceptual framework delineates an examination of Instagram data pertaining to video posts, encompassing features such as timestamp records, comments count, likes count, and total view count. By stratifying the posts across temporal segments and day-of-week delineations, valuable insights are obtained into the impact of posting chronometry upon the user content engagement. By the dataset amassed using Instagram scraping, a battery of machine learning classifiers is deployed to obtain discernible patterns and correlations, thereby providing predictive models adept at forecasting the most opportune time slots and days for obtaining heightened user engagement levels. The study reveals that Random Forest classifier produced the best results and had the highest accuracy.