Action ML Logo

Universal Recommender

The Universal Recommender (UR) is able to use almost anything known about the user to make recommendations better. It takes in user actions, profiles, context, and item metadata, all of which it uses to improve predictive quality. Other publicly available recommenders cannot do this, in fact the UR is one of a new generation of multimodal recommenders that have obsoleted most of the older technology.

Coming Soon: Launch on AWS


Set up a free consultation

Own your own data and your own future. You may need the technology but you don't have to pay for it by using the SaaS tech from the big names. The Open Source Universal Recommender gives you the flexibility you need.

Multiple data icon

Use all your data

The Universal Recommender can take in user actions, profile data, contextual information, virtually a user's entire clickstream, add that to content and meta-data to gain a deep understanding of user preferences, when others cannot.

Flexibility icon


The Universal Recommender has extremely flexible but simple queries for most any situation. It is also open source built so you can choose the hosting environment that works for you—start simple or customize any way that suits you.

Data and model icon

Own your data and model

Your data is your competitive advantage. With The Universal Recommender, you keep it private. Some SaaS vendors see this differently—not us. SaaS providers also charge for their operations—something you may not need.

Lock-in icon

No lock-in

It’s open source with no vendor lock-in. You can customize the code to fit your needs at anytime. You can't do that with AWS, Google, or Azure services.

Practical solutions, powerful examples.