Dynamic pricing engines
Real-time competitor price signals as input features for ML-based repricing models.
Product recommendation
Structured product attributes — title, category, brand, price — for collaborative filtering and content-based recommenders.
NLP & classification
Product titles and descriptions as training corpora for category classifiers, search ranking models, and sentiment pipelines.
Computer vision
Product image URLs with structured metadata for visual search, image quality scoring, and product deduplication models.
Anomaly detection
Price and stock time-series as input for anomaly detection — identify unusual market movements or competitor actions automatically.
Demand forecasting
Price and availability signals correlated with market demand — useful for inventory planning and supply chain models.
