Eray Korkmaz

I’m a Data Scientist with 3+ years of experience in fashion and apparel retail, focused on delivering production machine learning across computer vision, recommendation systems, and CRM analytics. I build SQL-first data pipelines and take ownership from data to deployment.

In my current role at Aydınlı Grup Hazır Giyim San. Tic. A.¸., I initiated image-based product recommendation deployments for major e-commerce brands, enabling discovery of over 750 previously overlooked product pairings. I also led YOLO, DINOv2, and SAM workflows for object detection and image segmentation, achieving %95+ improvement in image recognition accuracy.

I build scalable retrieval and forecasting solutions, including FAISS-based similarity search using image feature embeddings and low-latency retrieval with MongoDB plus a Redis caching layer. I support planning through weekly merchandise-level demand forecasting pipelines using NeuralForecast (NHITS/NBEATS) and LightGBM/CatBoost feature engineering.

Before that, as a CRM Analyst at Derimod Deri Pazarlama A.S¸., I owned CRM analytics end-to-end—RFM segmentation, campaign measurement, and KPI design—and designed A/B tests for SMS campaigns. I also mentor learners part-time at Miuul, guiding real-world Python/SQL, experimentation, and recommendation projects.