I am a Postdoctoral Researcher in Optimization Theory at Uppsala University, hosted by Prashant Singh. I work on deterministic and stochastic optimization algorithms with applications to machine learning and operations research. I have also been awarded a WASP postdoctoral scholarship from the Wallenberg Foundation for a two-year postdoctoral stay at Inria Paris, hosted by Francis Bach.

I completed my PhD in Mathematical Statistics at Umeå University in February 2026 as a WASP doctoral researcher, supervised by Alp Yurtsever. In my thesis, From Accelerated First-Order Methods to Structured Nonconvex Optimization: Analysis and Perspectives, I studied accelerated optimization methods through their continuous-time behaviour and structured nonconvex problems through difference-of-convex programming, with applications to machine learning. During my PhD, I was a research intern at Pendulum Systems and a visiting researcher at the University of Edinburgh and the Technical University of Munich.

Research interests: convex and nonconvex optimization, difference-of-convex programming, bilevel programming, accelerated algorithms, stochastic methods in machine learning, training algorithms, language models, reinforcement learning and bandits, inverse problems, and sparse representation theory.

Latest News

[September, 2026] Two of our papers were accepted at NeurIPS 2026: Tight Generalization Bounds for Noiseless Inverse Optimization, with Pouria Fatemi, Suvrit Sra, and Peyman Mohajerin Esfahani; and The Multi-Block DC Function Class: Theory, Algorithms, and Applications, with Pouria Fatemi, Alp Yurtsever, and Suvrit Sra.

[August, 2026] I served as a reviewer for TMLR (Transactions on Machine Learning Research).

[July, 2026] I organized a session at EUROPT 2026 in Linz, Austria, and served as a reviewer for the journal Advances in Computational Mathematics.

[June, 2026] I gave a talk on Tight Generalization Bounds for Noiseless Inverse Optimization at the SIAM Conference on Optimization (OP26) in Edinburgh, UK.

[May, 2026] I presented The Multi-Block DC Function Class: Theory, Algorithms, and Applications at the ELLIIT Focus Period on Optimization for Learning at Lund University.

[April, 2026] I started as a Postdoctoral Researcher in Optimization Theory at Uppsala University, hosted by Prashant Singh.

[April, 2026] I was selected to serve as a reviewer for NeurIPS 2026.

[February, 2026] I successfully defended my PhD thesis, From Accelerated First-Order Methods to Structured Nonconvex Optimization: Analysis and Perspectives, at Umeå University. I am honored to have Yura Malitsky as my thesis opponent and Sebastian Pokutta, Niao He, and Patrik Ryden on my defense committee.

[February, 2026] I was selected to serve as a reviewer for ICML 2026.

[December, 2025] Our paper Randomized Block Coordinate DC Algorithm, with Paniz Halvachi, Suvrit Sra, and Alp Yurtsever, was published in the EURO Journal on Computational Optimization.

[November, 2025] I was selected to serve as a reviewer for ICLR 2026.

[September, 2025] Our paper Revisiting Frank-Wolfe for Structured Nonconvex Optimization, with Yikun Hou, Suvrit Sra, and Alp Yurtsever, was accepted at NeurIPS **. I also received a travel grant from the **Kempe Foundation to attend the conference.

[June, 2025] I was selected to serve as a reviewer for NeurIPS 2025. I also served as a reviewer for the Journal of Optimization Theory and Applications (JOTA).

[March, 2025] I started a research visit to the Technical University of Munich (March–May 2025), hosted by Suvrit Sra. During the visit, I finalized a project on blockwise difference-of-convex programming and started new projects on portfolio optimization, bilevel programming, and inverse optimization.

[February, 2025] I presented Block Coordinate DC Programming at the Umeå University PhD Conference.

[January, 2025] I am delighted to inform that I was selected to serve as a reviewer in ICML 2025.

[November, 2024] My journey in Pendulum Systems was amazing. During my internship, I had the honor to receive supervision from Vahid Rostami. Together with Vahid, Alp Yurtsever, and Suvrit Sra, we successfully implemented an end-to-end optimization scheme for more efficient resource allocation. Several tests on real datasets suggested potential future for such optimization techniques. As my internship comes to an end, I’d like to thank Pendulum Systems for all the great experiences. I’ll definitely miss working with such a professional team and the great leader, Benjamin Fels.

[October, 2024] My course on Transform Methods (Digital Signal Processing) for this semester will begin on half-speed from November 2024.

[September, 2024] I am delighted to inform that I was selected to serve as a reviewer in ICLR 2025.

[September, 2024] I am delighted to inform that I was selected to serve as a reviewer in AISTATS 2025.

[June, 2024] I joined EURO and EUROPT 2024 in Lund. There I gave a talk on a new conditional gradient type method called DC-Frank Wolfe.

[March, 2024] I am thrilled to announce that I started an internship program as Student Researcher in the R&D Laboratory of Pendulum Systems.

[January, 2024] I am delighted to inform that I was selected to serve as a reviewer in ICML 2024.

[December, 2023] I am thrilled to announce that I was granted travel funds from Wallenberg AI, Autonomous Systems and Software Program (WASP) funded by the Knut and Alice Wallenberg Foundation.

[December, 2023] I am thrilled to announce that I was granted travel funds from JC Kempe Memorial Scholarship Foundation.

[September, 2023] I am delighted to inform that I was selected to serve as a reviewer in ICLR 2024.

[September, 2023] My paper titled: A Variational Perspective on High-Resolution ODEs was accepted for publication in NeurIPS 2023. This was a joint work with Konstantinos Zygalakis (Edinburgh University) and Alp Yurtsever (Umeå University).

[August, 2023] I am delighted to inform that I was selected to serve as a reviewer in NeurIPS 2023.

[August, 2023] I had a productive week at EUROPT 2023 Summer school in Budapest. There we had the opportunity of spending a day with Prof. Russell Luke and Prof. Coralia Cartis. I also presented my recent research entitled “A Variational Perspective on High-Resolution ODEs” in the EUROPT 2023 workshop.

[May, 2023] I started my recent project on Probabilistic Circuits (PCs). The aim is to use PCs as generative models which guarantee exact inference. Having priors generated from the PC, we leverage a Bayesian framework to recover noisy images.

[May, 2023] During my visit to the UK I will also visit the department of Mathematics in Oxford University and the Imperial College London. I will visit Prof. Jared Tanner in Oxford University and Prof. Tolga Birdal in Imperial College London.

[February, 2023] I will go on a research visit to Edinburgh University. There I will be working on the mysterious acceleration phenomenon in first-order methods together with Konstantinos Zygalakis.