Neural networks and machine learning
Artificial neural networks are computer algorithms mimicking the operating mode of our brain. These neural networks belong to the vast field of machine learning and have been employed for a wide range of applications.
Designing new materials and molecules—whether for medicines, electronics, or solar energy—requires understanding how they behave at the atomic level. This endeavor includes predicting how molecules pack into crystals or how they respond to heat or light, which can trigger chemical changes. Traditionally, simulating these properties with quantum mechanics is extremely accurate but also painfully slow. Our research uses machine learning to speed up these predictions by teaching algorithms to recognize patterns in outputs from quantum chemistry calculations—e.g. how molecules arrange themselves in crystals or how they evolve after absorbing light. These AI models learn from quantum-accurate data but run much faster, making it possible to explore new compounds and light-driven processes at a much larger scale. The goal is to bring powerful simulations closer to real-time decision-making in chemistry and materials science.
Creating effective visual content—whether for presentations, social media, or branding—requires both design skill and time. The goal of our research is to make high-quality graphic design more accessible by using machine learning to assist and automate parts of the creative process. This includes generating images, layouts, and text based on simple descriptions, making it easier for users to go from idea to finished design. We use large language models and image generation tools, such as diffusion models, to power features that can suggest layouts, rephrase text, edit images, or create new visuals from scratch. These AI systems are trained on vast examples of design work and can produce results in seconds, helping individuals and teams create polished, professional content—even without prior design experience.
Laser - molecule interactions
We study how molecules behave when they’re exposed to extremely short and intense laser pulses—like using a high-speed camera made of light to capture chemical reactions in action. These laser flashes last only a few femtoseconds (a femtosecond is one millionth of a billionth of a second, or 10⁻¹⁵ seconds), which is fast enough to track how atoms and electrons move during a reaction. Using quantum physics and computer simulations, we model how the laser field (see e.g. the Wigner representation of a third-order chirped pulse on the left) interacts with a molecule—how it can cause bonds to break, form, or rearrange. This helps us understand the fundamental steps of chemical change as they unfold in real time.