Onur Dogan

I am working as Machine Learning Engineer at Lifemote, designing and developing ML solutions for telecommunications analytics — including multimodal AI for device identification, production traffic-classification models, and LLM agents for anomaly explanation.

For my master's thesis, I worked on CoFine: Hierarchy-Aware Semi-Supervised Learning for Fine-Grained Behavioral Classification, under supervision of Asst.Prof. Oznur Tastan & Mehmet Keles from Johns Hopkins University.

Apart from my thesis, I am researching NLP applications in Computational Biology, focusing on Sequential Models, Protein Language Models (ESM2-Language models for protein sequences), Interpretability in NLP & LLMs. My intent is to develop more robust and interpretable machine learning systems to achieve fully intelligent systems that go beyond the data.

Prior to start pursuing MSc, I was member of Computational Biology Group and lucky to be supervised by Prof. Hilal Kazan. I worked on the development of computational methods for analyzing RNA-seq and scRNA-seq datasets under the supervision of Prof. Hilal Kazan.

On the industry side, I worked as a Computer Vision Engineer at Arioviz (Germany), developing a real-time segmentation model based on FastInst and SAM2 for robotic perception systems. I also had an amazing chance to work with Can Kavaklioglu as Junior ML Engineer at GrainFox, and I worked at John Snow Labs as a Junior Data Scientist and at Turkish Aerospace Industries as a Software Engineer Intern.

 /  CV  /   / 

Happy to share my latest CV on request.

Website template credits to Jon Barron.