System

Recosphere is a recommendation engine that combines the complexity of human character traits and the broad variety of human buying preferences in an optimal way with e-commerce products.

The system consists of the recommendation algorithm, a graph database, a data enrichment instance and the frontend.

Graph

The Graph database Neo4j is the heart of Recosphere, allowing you to store your product data connected to what you know about your customers.

By traversing the graph, Recosphere can evaluate an unprecedented number of criteria in realtime to recommend ideal products.

Intelligenz

Recosphere is a learning system: It calculates recommendations in just milliseconds, based on individual user preferences, while constantly learning from user feedback to improve the perceived quality of the recommendations.

Thresholds and quality grades protect the system for unbalanced changes.

Analytics

Recosphere’s analytic tools provide realtime tracking amongst other classic e-commerce evaluations. User interactions can be tracked in detail and used for funnel analyses combined with other reporting tools.

Valuable bases for decision-making and new learnings about your customers are the benefits of this system.

Modular

Different scenarios need an individual balancing of the algorithm.

Recosphere’s  modular system is precisely build for that: each module contains specific logic and separate weights that are used for algorithm scoring.

All modules benefit from continuous development.

API

Thanks to the open platform-independent architecture of the system, Recosphere can be customized for any individual needs.

Different data sources are easily connected and the recommendation results can be displayed in a personal frontend or via javascript plugin integrated into existing shops.