Bugra Veysel Yildiz

Responsible for designing and implementing a large-scale article recommendation system for Scopus, covering personalized and non-personalized retrieval. Work included full-cycle ML development such as data exploration, feature engineering, model training, and evaluation. Improved collaborative filtering models and integrated citation trends with regression and classification techniques for auxiliary prediction tasks.

Conducted applied machine learning research on multimodal gaming performance data. Processed telemetry and behavioral data to develop feature representations for predictive modeling and sentiment analysis. Designed ML pipelines for emotion recognition from audio.

Delivered AI/ML solutions for clients by transforming complex business problems into data-driven pipelines. Built prototypes and experimental models across NLP and analytics use cases, supporting model lifecycle management and providing feasibility advice to stakeholders.