Projects

AI security and machine learning

AI Firewall

I developed compact language models to detect malicious and policy-violating inputs at the application boundary. The project covers domain-specific tokenization, model training, quantization, and inference deployment.

Data Synthesis

I lead a team of four PhD researchers studying controllable synthetic data for AI security applications. Our work pairs a diffusion language model generator with an encoder-only discriminator to steer generated data toward target styles.

Computer networks

Network Architecture for 6G Networks

My doctoral research explored adaptable network architecture and protocols for delay-sensitive communication. It included work on transport protocol design, delay-guaranteed routing, and related scheduling methods.