The Twin Celebrity app is a facial recognition application that matches users with their celebrity look-alikes using advanced technology, including facial recognition embeddings and vector search powered by Qdrant. The app's architecture involves multiple tools, such as a ResNet-based FaceNet model for generating face recognition embeddings and ZenML for data pipeline orchestration, while Streamlit serves as the front-end interface. The offline phase involves preparing a dataset of celebrity images, aligning faces with MTCNN, and generating embeddings stored in Qdrant Cloud, while the online phase captures user selfies and queries Qdrant for the closest matches. Developed by Miguel Otero Pedrido, the app showcases the capabilities of modern vector databases and embedding-based applications, with potential real-world uses in security, healthcare, and retail beyond entertainment. The project emphasizes scalability and optimization strategies, such as quantization to handle large datasets, and suggests future enhancements like mobile-friendly deployment and real-time video input.