aurora-recommender-v2
Session-based recommender for the Aurora storefront.
Model Details
Graph neural network model that predicts next-item engagement from user session sequences.
Intended Use
Powers personalized product recommendations on the Aurora storefront.
Limitations and Bias
Recommendations may reflect popularity bias. Requires session context to be effective.
Training Data
Trained on Aurora clickstream data (anonymized) from 2024-2025.
Remediation
Missing license declaration.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support