Panagiotis Kyriakidis

I am a data scientist and ML engineer with hands-on experience in NLP, predictive modeling, customer segmentation, and semantic search. I build production-oriented systems using Python, transformers, FAISS, PyTorch/TensorFlow, FastAPI and Docker to deliver low-latency, scalable solutions.

My projects include a Movie Semantic Search Engine handling 800K+ items with sentence-transformers, FAISS and GPU-optimized cross-encoder re-ranking, a fake-news detection pipeline using BERT/LSTM achieving up to 78.33% accuracy across datasets, and customer segmentation plus predictive analytics with actionable visualizations in Power BI.

I combine strong analytical thinking, feature engineering and evaluation expertise with clear communication and a results-driven mindset, and I seek roles where I can apply modern NLP and ML techniques to solve real-world problems and help teams deliver measurable impact.