Users prefer options but too many can lead to analysis paralysis. Recommender systems help businesses offer plenty of options without overwhelming users. Content-based filtering is a technique that uses machine learning to suggest items based on their features and user preferences. It analyzes item features and user behavior to build a user profile that the system can match with new items. The quality and breadth of metadata are crucial for content-based filtering, as it often relies on these factors. However, it has limitations, such as struggling in domains with complex or unstructured content. Collaborative filtering methods, on the other hand, use user interactions to make recommendations. Hybrid recommender systems combine content-based and collaborative filtering techniques to balance their advantages and disadvantages. RedisVL can be used to build a content-based recommendation system by generating semantic embeddings for each movie's title, description, and keywords, and then using vector similarity search to find semantically similar movies. The system can also be expanded to include filters for specific fields set up in the schema.