A video understanding AI available as a cloud-based video recognition API that can effectively highlight and interpret concepts from any video stream. It creates scene level, time-coded metadata, providing all the needed content data through a single ingest point. Detect and identify faces, tag visual objects, places and activities, recognize sounds and speech, flag and filter inappropriate content and categorize content automatically. To use, upload videos to recognition via API or online tool. The AI analyzes the videos to get time-coded metadata with the most important concepts in the videos. Finally, integrate and use the content data in applications and workflows.