"I interned at CITIC (Center for Research in Information and Communication Technologies) at the Universidade da Coruña in Spain. CITIC is a leading research institution focused on AI, cybersecurity, high-performance computing, and intelligent networks. My project was centered around foundational AI models for cybersecurity. I specifically worked on applying Transformer-based architectures to passive OS fingerprinting using network traffic. Over the summer, I worked on fine-tuning ET-BERT, a pre-trained Transformer model designed for classifying network traffic. I built a dataset from scratch by capturing packet data (pcap files) from multiple operating systems—Linux, macOS, and Windows—using virtual machines. I then preprocessed the data to match the model’s expected format and fine-tuned ET-BERT for the OS classification task. All of this was to provide insight to the research group on the strengths and weaknesses of ET-BERT as a generalizable network traffic analysis model."
