Efe Parlak Email & Phone Number

Worms, Rheinland-Pfalz, Almanya

Developed a comprehensive system designed to detect identity fraud.- Created synthetic datasets for training a deep learning model.- Implemented various features, including optical character recognition (OCR) matching, template matching, selfie comparison, behavioral randomness detection, DOVID detection, and integrity checks for identification documents.- Prepared a FastAPI script for integrating the system into a mobile application.- To enhance system usability and ensure continuous operation, Dockerization was employed to automate the necessary setup, downloads, and file transfers.- The system’s primary advantage was its ability to generalize methods for integrating with identification documents from various countries.Conducted research and testing in the field of Face Anti-Spoofing (FAS). Contributed to data preprocessing stages for FAS datasets.Currently, I am engaged in developing a system capable of detecting deepfake videos and credit scoring.

Türkiye

- Developed data-driven solutions to address customer problems.- Identified data sources and accessed this data either via API or manually, performed data preprocessing, and subsequently predicted future prices using models built with LSTM and LightGBM.- Automated the extraction of customer requests from internal emails through text analysis to accelerate responses to customers.

Antalya, Türkiye

- Optimized stock management applications using LSTM- Implemented Association Rule Learning (ARL) for enhanced warehouse design optimization.- Extracted insights from customer datasets for informed decision-making.

İstanbul, Türkiye

- Forecasted maintenance and breakdown periods for manufacturing machinery.- Generated insights on machinery performance through exploratory data analysis using Python.