Yusuf Baysal

I’m an AI Researcher focused on ensuring AI system data quality and usability through careful research, testing, and evaluation. I thrive on turning feedback into continuous improvements for machine learning outputs.

My work connects AI research with practical user-centered design—optimizing AI model performance, strengthening natural language understanding, and building context-aware AI solutions. I’m especially interested in how reliable evaluation feedback loops can raise real-world system quality.

In my Master’s Capstone, I developed a BERT-based NLP model for sentiment analysis on Turkish movie reviews, including dataset preprocessing, tuning hyperparameters, and classifying reviews as positive or negative. The project strengthened my applied approach to NLP and text classification.

Through hands-on experience across prompt engineering, LLM development and fine-tuning, and generative AI applications, I aim to deliver high-quality AI systems. I bring research methodologies and A/B testing thinking to every iteration, from scenario design to model optimization.