Path to Science
Shaping the future of AI through mathematical rigor and algorithmic mastery.
I am a research scientist working at the intersection of stochastic analysis, control theory, dynamical systems, and generative artificial intelligence.
My goal is to design the next generation of generative models — systems that are not only predictive, but mathematically grounded, robust to uncertainty, and consistent with physical principles.
Technical Research Pillars
PI-DEs Analysis and Control
Development of the theoretical framework for evolutionary partial differential-integral equations, focusing on existence, stability, and controllability of complex systems.
Neural Networks for Evolutionary Equations
Leveraging deep learning architectures to approximate and solve high-dimensional dynamical systems, with applications to scientific computing and physics-informed models.
Optimal Transport & Flow Matching
Designing principled generative models based on optimal transport theory, diffusion processes, and flow matching, with a focus on stability, efficiency, and interpretability.
Graph Neural Networks and Applications
Modeling structured data through graph-based learning, with applications in networks, scientific data, and relational systems.
Kernel Methods in Machine Learning
Exploring kernel-based approaches for high-dimensional inference, bridging statistical learning theory with modern AI systems.
Open Source Contribution
SpectralGen
A research-driven Python library for spectral graph analysis and physics-informed generative modeling, designed to bridge theory and scalable AI systems.
Latest Trajectory
- Jan 2026 — Conditional Sampling for Deep Generative Models
- Ongoing — Integration of optimal transport, flow matching, and stochastic control in generative AI
- Focus — Unified frameworks for learning, dynamics, and uncertainty quantification