Data & Software Engineer · Systems Builder · Co-Founder & Tech Lead at @Xaol-Studio
Informatics Undergraduate at UNAM (FES Cuautitlán) · Google Student Ambassador 2026

I engineer high-performance data pipelines, embedded OLAP engines, and deterministic software systems. My technical foundation combines computer science at UNAM with practical experience in municipal public accounting and fiscal auditing.

Currently, I specialize in Data Engineering & Applied Analytics, designing scalable data lakes, zero-copy columnar ingestion pipelines (Apache Arrow + DuckDB + Parquet), and econometric time-series models for Mexican financial and retail markets.

• Data Engineering & FinTech Core: Building high-throughput ETL engines for Mexican SAT CFDI 4.0 XMLs, Banxico macroeconomic series, and transactional PyME retail datasets with zero external runtime expenses ($0 budget architectures).
• Co-Founder & Technical Lead at XAOL Software Studio: Architecting custom web platforms, relational inventory/POS engines, and compliance tools for Mexican businesses.
• Google Student Ambassador 2026: Organizing technical workshops, developer conferences, and AI/data initiatives across UNAM campus.
• Engineering Principles: Columnar-first storage (Parquet/Arrow); embedded analytical SQL (DuckDB); zero-knowledge client-side computation; strict automated test coverage across all releases.

Data Engineering:    DuckDB (OLAP) · Apache Arrow (PyArrow) · Parquet · Hive Partitioning · Streaming XML (lxml)
Data Science:        Pandas · NumPy · RFM Segmentation · ABC Pareto Analysis · Market Basket (Apriori) · Econometrics
Languages:           Python 3.10+ · TypeScript · SQL (PostgreSQL, SQLite, DuckDB) · Go · Java · PHP
Web & Mobile:        Next.js 15/16 (App Router) · React 19 · React Native / Expo (SDK 52) · Tailwind CSS
State & Storage:     Zustand · MMKV · IndexedDB · Web Workers · Zero-Knowledge Local-First
Systems & DevOps:    GitHub Actions CI/CD · Docker · Linux/Bash · Git · Pytest · Vitest
Domain Expertise:    Mexican SAT CFDI 4.0 / Anexo 20 · SAT Art. 69-B EFOS Detection · Banxico SIE API · IMSS 2026

High-Performance XML-to-Parquet ETL & DuckDB Analytics Engine · Python 3.12, DuckDB, Apache Arrow, Parquet, pytest

• Ingests bulk Mexican SAT CFDI 4.0 XML invoices via memory-constant chunked XML parsing (lxml).
• Batches records into zero-copy Apache Arrow tables, persisting into Hive-partitioned columnar Parquet lakes (year=YYYY/month=MM).
• Executes sub-second analytical SQL queries over hundreds of thousands of fiscal records using an embedded DuckDB OLAP engine.
• Instant $O(1)$ hash cross-referencing against the official Mexican SAT Article 69-B EFOS (shell company) blacklist.
• Links: Repository

Automated Macroeconomic ETL & Econometric Time-Series Engine · Python 3.12, DuckDB, Apache Arrow, Parquet, pytest

• Automated data pipeline extracting official Mexican central bank indicators (USD/MXN FIX, TIIE 28d, UDIS, INPC inflation) from the Banxico SIE API.
• Stores historical time-series in an embedded DuckDB data lake with partitioned Parquet backing.
• Codifies econometric routines: automated Fisher equation calculation (nominal TIIE vs. YoY INPC inflation to compute real interest rates) and cross-series Pearson correlation matrices.
• Links: Repository

Transactional Retail Data Science & Inventory Intelligence · Python 3.12, DuckDB, PyArrow, Parquet, pytest

• Production-grade data analytics engine converting raw transactional sales and stock histories into operational business intelligence for Mexican PyMEs.
• Quantile-based RFM (Recency, Frequency, Monetary) customer segmentation with behavioral cohort categorization (Champions, Loyal, At Risk, Hibernating).
• ABC Pareto (80/20) inventory classification, statistical Reorder Point (ROP) calculation, and Market Basket association rules (Apriori Support, Confidence, Lift).
• Links: Repository

Municipal Fiscal Audit Analytics & Automated PDF Reconciliation · Python, Streamlit, ReportLab, Pandas, pytest

• Born directly out of municipal government accounting experience: automates invoice reconciliation and municipal ledger matching.
• Generates structured, audit-ready compliance PDF reports programmatically with ReportLab.
• Comprehensive 42-test pytest suite guaranteeing calculation accuracy across tax brackets and ledger entries.
• Links: Repository

Co-Founder & Lead Engineer · Commercial Software & Business Platforms

• Independent software development studio building capture websites, relational inventory/POS engines, and fiscal processors for Mexican businesses.
• Architected the open-source pyme-manager-core engine: deterministic integer-cents financial math, SAT Anexo 20 tax reconciliation, and transactional POS cart state machines with native CI/CD.
• Links: Organization · Core Engine · Live Platform

In-Browser CFDI 4.0 Fiscal Auditor & SAT Art. 69-B Sentinel · TypeScript, Next.js 15, Web Workers, Vitest

• 100% client-side zero-knowledge architecture: parses thousands of XMLs and ZIP packages in memory without transmitting sensitive financial records to external servers.
• Screens suppliers in $O(1)$ time against the official Art. 69-B SAT blacklist (EFOS / simulated operations) and validates Anexo 20 tax formulas.
• Links: Repository · Live Demo

Mexican Payroll Reverse-Engineering (Net-to-Gross) & IMSS 2026 Labor Suite · Next.js 15, TypeScript, SheetJS, Vitest

• Inverts progressive Mexican income tax curves (ISR Art. 96) via sub-cent binary search convergence in < 25 iterations.
• Full compliance with 2026 progressive IMSS pension quotas and statutory LFT termination calculations.
• Links: Repository

Offline-First Mobile Productivity Suite · React Native, Expo SDK 52 (New Architecture), Zustand, MMKV

• Sub-millisecond synchronous storage via native MMKV bindings, preventing UI hitching or state desynchronization.
• Automated delivery pipeline: GitHub Actions builds, tests, signs, and releases production APKs automatically on version tag.
• Links: Repository · Web Demo

Combinatorial University Schedule Solver · TypeScript, React, pdf.js, Tesseract.js

• Backtracking algorithm with constraint satisfaction that evaluates thousands of course/group combinations against user-defined time windows and free-day rules.
• Includes client-side OCR and raw PDF timetable parsing running entirely in the browser.
• Links: Repository · Live Demo

• 100% Zero-Cost / High Efficiency: All data pipelines and analytics suites run on local, embedded engines (DuckDB, Parquet, PyArrow) without incurring expensive cloud or managed database bills.
• 100% Privacy by Design: Financial and fiscal tools operate client-side or on self-hosted infrastructure. Zero third-party telemetry on sensitive fiscal data.
• Automated Verification: Strict CI/CD workflows across projects enforcing multi-version testing, type checking, linting, and automated unit testing prior to merge.

I am open to technical discussions, freelance engineering contracts through XAOL Studio, and select Data Engineer / Data Analyst / Software Engineer roles (Remote or Hybrid in CDMX / State of Mexico).

• GitHub: @iamhuitron
• Studio: github.com/Xaol-Studio
• Email: ianhuitron0687@gmail.com
• LinkedIn: Ian Miguel Delgado Huitrón
• WhatsApp (Business / Studio): +52 593 126 9253