We present MeowID , a dual-expert retrieval system for individual cat identification from photographs. Existing metric-learning and animal re-identification methods enable embedding-based gallery matching, but their performance is often affected by variations in pose, viewpoint, and facial visibility in unconstrained images. MeowID integrates complementary whole-cat and facial recognition experts within a face-priority retrieval framework. When a usable aligned face is available, the face representation is augmented with a validation-guided whole-cat hint that provides complementary global appearance information; otherwise, the system falls back to whole-cat retrieval. This design exploits fine-grained facial cues while preserving robust identity recognition when facial evidence is unavailable. Route-specific embedding galleries further allow new individuals to be enrolled without retraining the recognition models. We also introduce Individual Cats in the Wild (ICW) , an open-source annotated dataset and benchmark for individual cat retrieval. Experiments against representative classification, metric-learning, and animal re-identification baselines show that MeowID achieves more accurate and robust cat retrieval.