Data science, machine learning and operations research for public policy and social good across Mexico and Latin America.
Escuela de Gobierno y Transformación Pública · Tecnológico de Monterrey
Our open-source work lives here — start with featurizer.
- Temporal ML for prioritizing and targeting public programs, with point-in-time correctness (no label leakage).
- Optimization / operations research for logistics, territorial coverage and resource allocation.
- Computer vision and applied AI.
- Training and capacity transfer to government teams.
Guiding principle: objective functions encode political choices. We document what is optimized, for whom, and the trade-offs between efficiency and equity.
| Project | Description |
|---|---|
🧩 featurizer · Apache-2.0 |
Automated feature engineering for temporal data, PostgreSQL-native: Deep Feature Synthesis compiled to pure SQL. |
| 🧪 triage-pg | PostgreSQL-native, simplified fork of DSSG triage for temporal ML on tabular public-policy data (teaching + consulting). |
Logistics and operations · electoral processes and geospatial optimization · education policy · financial inclusion · public health · philanthropy analytics.
Most of this work uses partners' sensitive data and lives in private repositories.