Santiago Segovia
Recent graduate from University of Houston, with a strong background in mathematics and computer science. Looking for full time opportunities in Data Engineering or Software Engineering.
Origin
I have a strong passion for mathematics and computer science, and I enjoy expanding my knowledge in these fields, particularly in numerical analysis, machine learning, and data science. In my free time, I like to work on personal projects that allow me to showcase my skills, read books, watch movies, play video games, and football.
profile.details() →Theorems
Selected work
Exploring computation, mathematical structure, and patterns in data.
January 2025 - May 2025
CompletedInvestigated GPU acceleration for dense linear systems through CPU, naive CUDA, and optimized CUDA implementations of LU decomposition, achieving approximately 40–41× end-to-end speedup on a 5000 × 5000 benchmark.
CUDA / C++
Source codeJune 2025 - August 2025
CompletedDeveloped a Python information retrieval system for attribute- and opinion-focused product review search, comparing retrieval approaches through manual Precision@k evaluation and iterative query refinement.
Python
Source codeJune 2025 - August 2025
CompletedContributed to a full-stack volunteer management platform integrating role-based workflows, event coordination, profile-informed matching, assignment tracking, and reporting, with PHPUnit tests for profile validation.
PHP / JavaScript / HTML / CSS / SQL
Source codeAugust 2025 - December 2025
CompletedInvestigated random, banded, and block sparsity in MATLAB to examine how matrix structure affects computational cost and predictive accuracy in an arithmetic model motivated by encrypted machine learning inference.
MATLAB
Source codeMay 2026 - Present
CompletedDesigned and developed a Next.js and TypeScript portfolio combining reusable project data, dedicated technical case studies, interactive mathematical graphics, responsive layouts, and a Resend-powered contact workflow.
TypeScript / JavaScript / HTML / CSS / React / Next.js / Tailwind CSS
August 2026 - September 2026
CompletedAnalyzed personal Super Smash Bros. Ultimate match records using descriptive statistics, matchup heatmaps, K-means, XGBoost, and SHAP to examine character performance and explain patterns associated with recorded outcomes.
Python / pandas / scikit-learn / XGBoost / SHAP
Source codeMathematical sketches are illustrative.
Connect the Points