Tiago Cogumbreiro

Tiago Cogumbreiro is an Assistant Professor at the University of Massachusetts Boston, where he has been a faculty member since Fall 2018. His research focuses on advancing the foundations of parallel programming through rigorous quality assurance of languages and runtimes.

Dr. Cogumbreiro received his PhD from the University of Lisbon (ULisboa) in March 2015 under Francisco Martins, where he developed techniques to handle barrier deadlocks including the Armus runtime verification tool. His educational background includes a B.Sc. from Universidade dos Açores, followed by research assistant work at Imperial College London (supervised by Nobuko Yoshida) and postdoctoral research at Georgia Tech and Rice University (supervised by Vivek Sarkar).

As an expert in formal methods for high-performance computing, Dr. Cogumbreiro's work centers on detecting concurrency errors in parallel programs with special emphasis on GPU systems. His research spans theoretical contributions to deadlock avoidance policies and practical applications of Coq and Why3 for certified algorithms. Recent work demonstrates sophisticated approaches to static analysis of data-races in GPU programming, addressing fundamental challenges in parallel system reliability.

His publication pattern reveals a consistent trajectory from foundational work on futures-based deadlock avoidance (2017) through behavioral type systems (2019) to current innovations in GPU program analysis (2023-2024). These contributions form a cohesive research program focused on mathematically rigorous approaches to parallel system correctness.

Dr. Cogumbreiro actively contributes to the programming languages community through committee service at major conferences including PLDI (2020 Artifact Evaluation), SPLASH (2025 OOPSLA Review Committee), and PPoPP. His GitHub activity shows ongoing development of formal verification tools, particularly Coq-based projects like gorn-coq and habanero-coq that implement his theoretical contributions.