AI fashion assistant that analyses user style and outfit compatibility to deliver personalised clothing recommendations, integrated with the ASOS API for real-world product matches. Achieved 70–95% accuracy using 9 CNNs and ResNet-50 transfer learning, earning praise from industry experts for its performance and deployment.
Developed as capstone project for Le Wagon Data Science Bootcamp (Mexico) - a 10-week intensive course on Machine Learning, Deep Learning, Data Engineering & Data Analysis.
📹 Watch Demo Video
🚀 Download Project Pitch (PDF)
- Data & Preprocessing: NumPy, SciPy, OpenCV, Pillow
- Machine Learning & Deep Learning: TensorFlow, Keras (CNN classifiers, ResNet-50 fine-tuning), scikit-learn
- Backend: FastAPI (with Uvicorn + Gunicorn for serving)
- Deployment: Google Cloud Run (containerised API), Google Cloud Storage
