Daniel E.
González Limas

Mathematics & Data Science

Tecnológico de Monterrey · on exchange at Wilfrid Laurier University

About

I'm a Data Science and Mathematics Engineering student at Tecnológico de Monterrey, currently on academic exchange at Wilfrid Laurier University in Waterloo, Ontario. My work sits where rigorous mathematics meets applied problems: over the past two years I've built models for industrial and non-profit partners while training in the theory underneath them.

I'm preparing applications for master's programs in mathematics, with interests in analysis, algebra, optimization, and topological data analysis. Alongside that I write and record El Cuaderno de Euler, a Spanish-language mathematics channel.

Selected Work

Topological Data Analysis for Stockout Prediction

Ternium

A study of whether persistent homology adds predictive signal to raw-material stockout forecasting at one of Latin America's largest steel producers. I built topological feature pipelines over multivariate supply time series and tested them against a strong tabular baseline across six independent evaluation protocols. The result was negative and I reported it as such: the baseline reached AUC 0.84, and topological features failed to improve on it under any of the six evaluations. Establishing that cleanly — rather than tuning until the desired result appeared — was the substance of the work.

  • Python
  • giotto-tda
  • scikit-learn
  • XGBoost
  • pandas

Planogram Optimization with a Genetic Algorithm

OXXO

Shelf-space allocation for Mexico's largest convenience-store chain. The problem is combinatorial: assign products to shelf positions under category-adjacency, facing-count and physical-capacity constraints while maximizing expected margin. I modeled it as a constrained permutation problem and solved it with a genetic algorithm using custom crossover and repair operators that keep every candidate solution feasible, rather than penalizing infeasibility after the fact.

  • Python
  • DEAP
  • NumPy
  • pandas
  • matplotlib

Secure Identity and Access System

Casa Monarca

An identity and access-management system for a migrant shelter in Monterrey that handles sensitive personal records. Role-based access control separates coordinators, staff and volunteers; two-factor authentication supports both TOTP and WebAuthn; credentials are stored hashed rather than in plaintext; and every read or write against a beneficiary record is committed to an append-only audit log, so access to protected data is reconstructible after the fact.

  • Python
  • WebAuthn
  • TOTP
  • RBAC
  • PostgreSQL