Agoda, the online travel platform, deployed a multimodal AI system that fuses image recognition and text-analysis models to understand user intent across reviews, photos, and property descriptions. The system reduces cold-start friction in personalized recommendations for new users and properties.
For enterprise AI teams building recommendation systems, Agoda's infrastructure demonstrates practical value of multimodal fusion in high-cardinality catalog problems. Cross-modal ranking and retrieval at scale require careful model orchestration and infrastructure investment, but user engagement gains justify the complexity.