Content Discovery 

About the team

The Content Discovery Research team presents content to users in the Showmax app so they always find something to watch and stay with Showmax. To recommend the most relevant content for each user, we process the user’s viewing habits and all kinds of asset metadata. These are used in our statistical and ML models, e.g. the Film2Vvec model.

Apart from recommendations, the CDR team feeds various metrics to our Content team (people that actually understand the African shows). The tech stack consists mostly of SQL, Python, and Pandas, but that’s not all – the Film2Vec model is done using TensorFlow and they also maintain a small Flask app that visualizes what we recommended to each user. Most of their work is about analyzing, collecting insights, and providing personalized or automated content recommendations.

The CDR team works a lot with Showmax CMS, Backend, and Content teams. Externally, the CDR team works with Recombee which provides some recommendations.