Whenever a user clicks on “next video”, a signal is sent to Raptor’s recommendation engine. The Raptor recommendation engine then starts the process of finding the most relevant video for the individual user on Nordic Hiit’s platforms.
The algorithm starts by separating the three different training levels that the individual user has chosen in their profile preference settings. This means that there are three different algorithm paths which are either beginner, intermediate, or advanced.
The recommendation engine then starts a data filtering process by looking at the individual user’s profile and based on the chosen equipment, training goal, and training days it will include and exclude videos from the video catalog containing 1600 videos. Likewise, the algorithm will make sure that the videos fit the current season, so the workout videos that fit into a winter environment do not get recommended in the summer months.
In the last step, the algorithm takes into consideration which workout videos you have watched recently. If you have just done a workout that trained your legs and abs, then the next video should not contain the same exercises.
Based on the above, the algorithm will find the most relevant video.
The algorithm starts this process every time a user clicks the “next video” button, to find the correct order of videos to show. All these steps in the algorithms happen within milliseconds, and therefore there won’t be any delay for the user.