HSAL

Data Scientist · Monterrey, Mexico

Hendrik Steven
Arias López

I turn data into decisions: route optimization, customer segmentation, automation with Python and APIs, and generative AI. I build and run production systems end-to-end, and measure them against KPIs and ROI.

Email me LinkedIn Download CV (PDF)
30–60 s → <2 sgarment previews, custom image compositor
1,600+automated tests on a solo-built product
197%projected ROI, customer-segmentation proposal
14.6×GPU speed-up, underwater vision pipeline

Experience

Aurvy

Founder & Lead Engineer

AI-native apparel store: a text prompt becomes a design, previewed on garments, paid for and shipped. Built and run solo, with AI coding agents as an accelerator.

  • Automated the end-to-end order flow by integrating the OpenAI image API, Stripe (checkout, tax, webhooks) and a print-on-demand fulfillment API, with idempotent processing and automatic refunds.
  • Designed the architecture and data model: FastAPI, PostgreSQL, Celery workers and Docker on DigitalOcean; Next.js/TypeScript on Vercel.
  • Cut garment previews from 30–60 s to under 2 s with a custom Python image compositor.
  • Quality and monitoring: 1,600+ pytest tests, Playwright E2E, load tests (p95 under 500 ms), Sentry/Prometheus, zero-downtime deploys.
  • Business operations: IMPI trademark filing, LFPDPPP compliance with automated data-rights (ARCO) flows, SAT/CFDI invoicing, Meta Ads.

Industrias Alimenticias ARIS

Data Science Intern

  • Optimized distribution routes with stochastic optimization to reduce transport distance and cost.
  • Built customer segmentation with k-means (k = 6), grouping clients into six segments to support commercial and operational decisions.

TCA Consulting

Data Science Consultant · industry-partner capstone

  • Led Deep Embedded Clustering model training for the client's customer segmentation on Snowflake; the proposal projected a 197% ROI.
  • Reproducible pipelines with Kedro, experiment tracking with MLflow and GPU training (CUDA).
  • Built stakeholder-facing analytics dashboards in Streamlit.

Tecnológico de Monterrey

Research Intern · Multi-Objective Optimization

  • Hyper-heuristic for Pareto subset selection; article under review.

Casa Monarca

Cybersecurity Specialist

  • Developed a Python desktop application and a React web application.
  • Implemented a deep-learning facial recognition pipeline (97% accuracy) and cryptography-based data protection.

Selected project

VANTEC — Autonomous Submarine Vision

Corrected underwater backscatter and color distortion so YOLO detection could run in real time, with a 14.6× speed-up from GPU acceleration (CUDA).

Technical skills

Languages
Python, SQL, TypeScript, R, C++, MATLAB
Analytics & ML
scikit-learn, PyTorch, TensorFlow/Keras, k-means clustering, stochastic & multi-objective optimization, OpenCV, MLflow, CUDA
AI & Automation
OpenAI API (LLMs), API and webhook integration, FastAPI, Celery, Claude Code
Data
PostgreSQL, Snowflake, Kedro pipelines, Azure
BI
Power BI, Streamlit
Engineering
Git, Docker, pytest, Playwright, Sentry, Prometheus, Vercel, DigitalOcean, Cloudflare

Education & more

Tecnológico de Monterrey

B.S. in Data Science & Mathematical Engineering · Jul 2025

Coursework: Multi-Objective Optimization, Nonlinear Optimization, Abstract Algebra, Cybersecurity, Quantum Computing.

Honors

  • Outstanding Student Award, 2024
  • 1st Place, ExpoIngeniería, 2024

Leadership

Director of Financial Research, IMEFU Campus CCM (2022–2023). Led nationwide workshops reaching 100+ students.

Certifications & languages

  • Certified Ethical Hacker (2019)
  • ITMO Data Science (2022)
  • Bloomberg (2018)
  • Spanish (native) · English (B2)

Let's talk.

Open to data science, optimization and AI engineering roles.

hendriksteven15@outlook.com LinkedIn Download CV