Pasquale De Rosa

Machine Learning Scientist/Engineer.

prof_pic.png

Hi, I’m Pasquale De Rosa.

I hold a PhD in Computer Science from the University of Neuchâtel, Switzerland, where my research bridged the gap between advanced machine learning theory and real-world system engineering.

🛠️ What I Do

  • Adversarial AI & Security: Building robust intrusion and malware detection systems. I developed PhishingHook, an AI framework utilizing multimodal deep learning to isolate behavioral anomalies and de-obfuscate malicious EVM bytecode.
  • Quantitative Analytics: Modeling high-dimensional financial time-series to capture cross-asset causal relationships and predict trends in highly volatile, noisy data environments.
  • Production MLOps: Testing the practical boundaries of AI systems. I explicitly benchmark the latency–cost trade-offs of modern serving frameworks (TorchServe, MLflow, BentoML) to engineer optimal cloud deployments.

🎸 Beyond the Code

When I’m not profiling pipelines or reading threat vectors, I’m usually deeply immersed in high-energy or highly technical hobbies. I am a passionate retro gamer, a comic book collector, a rock/metal enthusiast, and a beginner keyboardist. I am also an active advocate for open-source systems.

🔍 Explore My Work

This site serves as an interactive portfolio of my technical journey.

  • [Interactive CV / Resume] — Explore my end-to-end engineering track record.
  • [Selected Publications] — Read my peer-reviewed work across IEEE and ACM.
  • [Technical Blog] — Where I break down complex security, MLOps, and causal inference concepts into clean, accessible architecture notes.

Latest post

Latest publication

  1. ACCESS
    Seeing Through EVM Bytecode Obfuscation
    Pasquale De Rosa, Pascal Felber, and Valerio Schiavoni
    IEEE Access, 2026