← Home

SCIENTIFIC EVIDENCE

The work can be reviewed before it is commissioned.

Consulting is not sustained by generic promises. It is grounded in open research on longitudinal analytics, pathways, causality, graphs, artificial intelligence and complex systems.

Publications that make the method, criteria and results reviewable.

A selection of open works. Each record retains its nature as article, thesis, preprint or working paper.

2025

A leakage-aware data layer for student analytics: the CAPIRE framework for multilevel trajectory modeling

PREPRINT · arXiv
2025

The CAPIRE framework: a unified architecture for curriculum-aware, multilevel, causal and simulation-based modelling

PREPRINT · EdArXiv
2025

An agent-based simulation of regularity-driven student attrition

PREPRINT · arXiv
2023

College dropout factors: an analysis with LightGBM and Shapley's cooperative game theory

PREPRINT · arXiv
2023

Student delay, dropout and university curriculum in the civil engineering major at FACET-UNT

THESIS · Zenodo
2025

From linear risk to emergent harm: complexity as the missing core of AI governance

WHITE PAPER · Zenodo
WHAT THIS EVIDENCE SHOWS

Results are not merely observed: explanations are built for review.

The work documents a path from fragmented data to structural diagnosis, causal interpretation and intervention scenarios.

SCOPE AND CAUTION

The strongest empirical validation was developed in education.

Its transfer to other sectors is methodological: it identifies mechanisms, constraints and delayed effects where a dashboard is not enough.