Eylül Şahingöz

I am a bilingual (EN–TR) linguist with a BA in English Language and Literature and a strong foundation in computational linguistics. I focus on evaluating LLM outputs for fluency, accuracy, and cultural relevance, and I have run self-directed benchmarking projects across multiple models.

I leverage 5+ years of teaching and linguistic error-analysis experience to create reproducible annotation guidelines and high-quality labeled data. My projects include LLM comparisons (ChatGPT, Gemini, Claude), translation error analyses, and sentiment annotation for Turkish content.

I am passionate about ethical AI and human-in-the-loop systems, and I combine classroom-tested assessment techniques with practical prompt testing and refinement to improve AI-generated educational and technical materials.