At TU Dortmund University, I’m pursuing an MSc in Data Science, with coursework in advanced statistical learning, simulation-based inference, and statistical theory.
On Double-Response EAM, I developed a Racing Diffusion Model to simulate two-choice decision-making and double-response behavior. I also implemented Bayesian parameter estimation and evaluated it using parameter recovery, Simulation-Based Calibration, and posterior predictive checks.
I built AstroChat by fine-tuning T5 on a self-built Q&A dataset, reaching 87.4% response accuracy. The project placed in the top 10 among 200 entries at the university project showcase, where I presented it to industry specialists; the related research paper was published at ICICC 2024 and indexed in Scopus.
My other projects include a RASA chatbot with over 82.5% intent recognition accuracy, an image-captioning pipeline using EfficientNet and RoBERTa, and a BERT-based sentiment analysis model. As a Benefits Intern at VENSURE HR, I administered employee benefits programs and maintained benefits records in the HR database.

