A system for generating video clips includes a video processing system configured to receive computer vision (CV) output data, which includes optical character recognition (OCR) text of graphics from a video, time codes for the OCR text of the graphics, and ontology classifiers for the OCR text. The video processing system includes a file storage and database system configured to store the video, and to store the CV output data according to an ontology associated with the ontology classifiers. A CV moments module of the video processing system is configured to identify themes contained in the CV output data using the ontology classifiers, the ontology, and natural language processing (NLP), and to identify time codes associated with the identified themes. A video clip generator of the video processing system is configured to generate video clips of the video based on the identified time codes.