I am a post-doctoral researcher at EPFL, working with Prof. Ali H. Sayed. I received my PhD in Electrical and Computer Engineering from Carnegie Mellon University, where I was advised by Prof. Soummya Kar. Prior to that, I earned my Bachelor’s and Master’s degrees in Applied Mathematics from the Faculty of Sciences, University of Novi Sad.

I am on the academic job market for tenure-track faculty positions starting in Fall 2027.

Research Interests

My research spans stochastic optimization, machine learning, and multi-agent systems. I seek to understand what performance guarantees are achievable in learning systems under challenges such as heavy-tailed noise, data heterogeneity, and communication constraints, and how to design algorithms that attain these guarantees. I pursue this agenda in two main directions:

• Robust and reliable stochastic optimization - understanding and improving the performance of stochastic optimization methods beyond classical noise assumptions, with an emphasis on robustness and individual-run reliability.

• Collaborative learning over heterogeneous networks - understanding the benefits and limitations of collaboration among learning agents in the presence of data heterogeneity, communication constraints, and decentralized information exchange.

Together, these directions contribute to my long-term vision of developing theoretical foundations for efficient and reliable learning systems in which intelligent agents can adapt to their environment and autonomously decide when and with whom to collaborate.

For more details about my research and publications, see my Research page. If you have any questions about my research or otherwise, feel free to reach out.

News

• October 2026 - A new preprint is out. Title: Decentralized SGD under Heavy-Tailed Noise: Optimal Convergence Rates and the Role of Gradient Clipping.
• Jun 2026 - A new preprint is out. Title: A Unified Zeroth-Order Approach for Decentralized Minimax Optimization.
• May 2026 - Our paper Tight Long-Term Tail Decay of (Clipped) SGD in Non-Convex Optimization got accepted for publication at the 39th Conference on Learning Theory (COLT).
• April 2026 - Our paper High-probability Convergence Guarantees of Decentralized SGD got accepted for publication at the 43rd International Conference on Machine Learning (ICML).
• April 2026 - A new preprint is out. Title: High-Probability Convergence in Decentralized Stochastic Optimization with Gradient Tracking.