I'm a theory-oriented computer scientist interested in combinatorial structures and the hunt for tight bounds. My work involves constructing extremal examples, finding structural invariants, and studying when natural upper bounds are tight.
Currently, my research focuses on de Bruijn-type constructions and their interaction with the Burrows-Wheeler transform. I am especially interested in how cyclic combinatorial structure controls repetitiveness measures and the limits of compressed indexes.
More broadly, I like problems where discrete structure meets computational constraints: string algorithms, compressed indexes, extremal constructions, and the mathematics of efficient representation.
I work with Leandro Zatesko as part of the Theory Group at UFPR.
String algorithms, combinatorics on words, de Bruijn sequences, the Burrows-Wheeler transform, suffixient sets, repetitiveness measures, compressed indexes, and finite-field/LFSR constructions.
On the near-tightness of χ ≤ 2r: a general σ-ary construction and a binary case via LFSRs
Shows that the bound χ ≤ 2r is asymptotically tight, via a general σ-ary construction and a binary construction from LFSR-generated de Bruijn sequences.
One proof uses OEIS A073639 as a condition; the entry now links the paper.
Successor right-special strings with few Burrows-Wheeler Transform runs
Studies a rigid de Bruijn-type regime where every context has exactly two fixed successor extensions. Gives explicit low-run BWT constructions and closed-form χ/r ratios approaching 2.
Exploring graph representation strategies for text classification
Um relato da experiência de construir um sistema para integração de dados e criação de indicadores para um programa governamental
On the near-tightness of χ ≤ 2r
Repetitiveness measures in compressed string indexes: suffixient sets, BWT runs, and de Bruijn sequences. MSc dissertation, UFPR, expected 2026. Advisor: Leandro Miranda Zatesko.
east_shmup
Ongoing C++/Vulkan systems project: a small renderer/runtime for a Touhou-style bullet hell, mainly as a way to learn GPU-facing graphics programming from the ground up.
Ranked 1st among MSc proposals from UFPR's Department of Informatics.
Top graduating student of the UFPR Computer Science cohort.
Twice qualified and competed in the Brazilian regional phase of the ICPC-style SBC Programming Marathon.
Teaching assistant at UFPR for algorithms and graph theory, including guest lectures on max flow / min cut. Co-lectured an ICPC-style problem-solving course and coached a student team that qualified for the Brazilian national finals in 2024.