Moreton Capital Partners — Mexico City, México

Architecting multi-agent heuristic systems operating as non-linear layers atop production-level deterministic models to optimize systematic commodity strategies and signal blending. Formulating mathematically rigorous meta-models to map high-dimensional financial datasets, enforcing statistical constraints and regularizers to ensure structural stability and mitigate alpha decay. Bridging stochastic machine learning outputs with classical convex optimization frameworks, integrating agentic signals into multi-horizon risk-parity and regime-aware allocation models.

Working Paper

Formalizing an end-to-end, model-agnostic architectural blueprint for processing high-dimensional, unstructured financial text, ensuring structural invariance against shifting baseline language architectures. Designing a modular data engineering pipeline that decouples stochastic noise-filtering mechanics from downstream deep learning systems, optimizing input embedding spaces to ensure future-proof compatibility. Formulating a novel, domain-specific reward function and objective metric that maps linguistic signal extraction directly to portfolio optimization, independent of the underlying transformer or LLM paradigm.

Instituto Via Diseño — Querétaro, México

Architected a constraint satisfaction scheduling system for a 150+ user campus environment, formalizing combinatorial resource allocation as a CSP and applying backtracking search with constraint propagation to guarantee feasibility under competing institutional constraints. Engineered PostgreSQL indexing strategies that reduced query latency by 40% under high-cardinality constraint evaluations, and surfaced scheduling state through a real-time React visualization layer.

Developed a two-stage backtesting engine for 10 commodity sectors over 162 weekly periods using SLSQP min-variance optimization, achieving max 7.7% annual volatility. Enforced KKT optimality for stable weight constraints and mplemented Ledoit-Wolf covariance shrinkage and a Stage A signal blending / Stage B cross-sector allocation pipeline that reduced Max Drawdown by 15%.

Syngenta @ START Hack — Top 5 Finalist (30 teams)

Engineered a real-time computer vision system to monitor crop health in Martian greenhouses, utilizing AWS SageMaker for autonomous stress detection. Developed a multi-modal data pipeline to process environmental telemetry and spectral imagery, enabling automated nutrient deficiency diagnosis in extraterrestrial environments.

Originally designed as a predictive ML model for drought patterns in CDMX using hydrological data. Pivoted to a crowdsourced reporting system to bridge data gaps. Designed the data ingestion pipeline to validate user reports against historical meteorological norms.

@ NASA Space Apps Hack

Engineered a LightGBM-based predictive model leveraging multi-spectral satellite imagery (GEE) and NASA POWER meteorological data. Conducted rigorous dimensionality reduction across 44 distinct ecological indicators to classify global bloom patterns with high precision (AUC/F1 validation).

Simulated autonomous agent behavior using Python (Mesa) and Unity to analyze strategic decision-making dynamics. Designed reward-based optimization functions within constrained state spaces, applying Monte Carlo sampling to identify emergent Nash equilibrium patterns.