I am building a recommendation engine as a Redis module. It uses Jaccardian similarity and memory-based collaborative filtering.

Any early feedback (constructive or destructive) is welcome.