Artificial Intelligence (AI) and Generative AI (GenAI) are transforming every industry. Our main goal is to ensure that AI technologies are not only powerful but secure, reliable, and aligned with the latest ethical, legal, and technological guidelines.

In that sense, a key focus of our research is making AI systems more secure during inference. We develop techniques that monitor and analyse input data in real time, comparing it with training data to detect anomalies, preventing threats such as data and prompt poisoning and mitigate issues like data drift.

We are also exploring the potential of Large Language Models (LLMs) and GenAI agents across diverse domains. Our work spans from translating HPC programming languages to enabling natural-language question answering across various use-cases, with a focus on applying these technologies to address concrete, real-world challenges.

Finally, our research looks at how data is generated and improved. We provide techniques for data augmentation as well as creation of advanced data pipelines, supporting robust workflows that prepare high-quality data for AI training.