Salva is a third-year PhD candidate at UC San Diego, working under the guidance of Prof. Rose Yu and Prof. Duncan Watson-Parris. His research focuses on developing and applying advanced machine learning (ML) techniques to address critical real-world challenges in climate modeling, weather forecasting, and sustainability. Salva's expertise lies in self-supervised learning, large-scale forecasting, and generative modeling. His work has led to the creation of novel, efficient generative AI algorithms specifically designed for temporal problems. A notable achievement is the development of an ML-based climate model emulator that significantly outperforms conventional physics-based models in terms of speed and cost-effectiveness while achieving similar simulation accuracy.

Salva's research has garnered recognition in the academic community, with multiple publications at NeurIPS, a premier ML conference. Additionally, his work received a best paper award at the ICML workshop on ML for Earth System Modeling. His collaborative spirit has led to internships and partnerships with leading institutions, including NVIDIA, Allen Institute for AI (AI2), and Mila - Quebec AI Institute.