It R&D Project Engineer Intern
Migros Genel Müdürlük, Atasehir, Istanbul
• Led the development of the Dash Button project, which allowed customers to order pre-selected products with a single click, delivering them within twenty minutes. The project involved programming the ESP8266 microcontroller to securely connect to the Wi-Fi network and trigger an API call that communicated with the backend system to process orders.• Integrated computer vision capabilities into an autonomous in-store robot using TensorFlow, enhancing its ability to identify and categorize products on shelves. I trained the neural network models on a custom dataset of product images, which allowed the robot to recognize and scan barcodes, detect misplaced items and monitor stock levels. This solution involved optimizing the model for real-time inference, ensuring it could run efficiently on the robot's onboard hardware while maintaining accuracy. Additionally, I worked on streamlining the data pipeline for image processing and implemented techniques for faster object detection.• Developed an Android application to detect pricing errors on product labels, streamlining the workflow for store employees. The app utilized OCR technology to scan product labels and match the extracted information against the database of current prices. I designed the app’s architecture to efficiently process large volumes of data, with features that allowed users to flag discrepancies and submit error reports directly to the central system. The app also supported batch scanning and offline functionality, ensuring it could be used in low-connectivity environments.
