Carlos León

Carlos León

Data Engineer & builder

Data Engineer atAstrafy

I build data platforms by day and products by night.

Building DataForgeMadrid, Spain

About

I'm a Data Engineer at Astrafy in Madrid, where I build data platforms and data products on Google Cloud with dbt, BigQuery, Airflow, and Terraform. I'm Google Cloud certified as a Professional Data Engineer and Professional Cloud Architect. Before that I did BI for the Spanish Ministry of Defense, SQL-heavy consulting at Minsait, and machine learning at Red Eléctrica's innovation lab.

I studied Mathematics, and I still approach problems the same way: understand the structure first, then build. Lately that means building a lot outside work too: a certification prep SaaS, the ML behind a hackathon financial-health engine, a personal finance dashboard, and a website for a small fashion brand.

I'm most drawn to where engineering meets product and startups: figuring out what's worth building and getting it working end to end. Outside of that, it's music, economics, literature, and too many essays.

Now

Building DataForge

Taking my certification prep app from a personal study tool to a real product: more certifications, a pipeline to generate and validate questions from official docs, and the pieces needed to open it to the public.

Builds

Products, side projects, and client work. Things I've taken from an idea to something people can use.

  • DataForge

    Live · free beta

    Certification prep SaaS with practice mode, timed mocks, and a gamified study loop.

    • React
    • Vite
    • Supabase
    • PostgreSQL
    • Vitest
    Video
  • X-Ray

    Live demo

    Financial-health engine that scores 1,286 companies from their bank trail. HackSpain 2026, Embat track.

    • Python
    • FastAPI
    • DuckDB
    • scikit-learn
    • Next.js
    Video
  • Wealth Tracker

    Live · public demo

    Self-hostable net-worth dashboard with live prices and natural-language journaling.

    • Next.js
    • TypeScript
    • Supabase
    • PostgreSQL
    • LLM
    Video
  • irisverd

    Live · shop coming soon

    Editorial website for a handmade clothing atelier, built for a real client.

    • Next.js
    • TypeScript
    • Tailwind CSS
    • Vercel

Data products

End-to-end data and ML pipelines, built the way I'd build them at work: reproducible, tested, and honest about results.

  • Madrid NO₂ Forecasting

    Leakage-safe forecasting of station-level NO₂ in Madrid at 1, 24, and 72 hours ahead.

    • Python
    • PostgreSQL
    • dbt
    • XGBoost
    • pytest
  • Chicago Taxi Trips vs Weather

    End-to-end GCP pipeline combining Chicago taxi trips and weather signals.

    • Google Cloud
    • BigQuery
    • dbt
    • Terraform
    • Cloud Functions
View all projects

Experience

  1. Sep 2025 - Present

    Data Engineer · Astrafy

    Building data platforms and data products on Google Cloud for finance and marketing teams, from ingestion and dbt modeling to CI/CD, infrastructure as code, and self-serve analytics.

    • Python
    • SQL
    • dbt
    • GCP
    • BigQuery
    • Airflow
    • Terraform
    • Looker
    • Lightdash
  2. Oct 2024 - Sep 2025

    BI Engineer · SII Group Spain

    Developed data engineering and BI solutions for the Ministry of Defense, improving planning, reporting, and visibility across financial and operational data.

    • Python
    • SQL
    • Power BI
    • Bash
    • Docker
    • Elasticsearch
  3. Jul 2023 - Jan 2024

    Data Consultant · Minsait

    Worked on SQL-heavy data solutions, partnering with analysts and technical teams to improve performance, reliability, and process quality.

    • SQL
    • PostgreSQL
    • ETL
  4. Apr 2023 - Jul 2023

    Data Science Intern R&D · Red Eléctrica

    Worked at the intersection of analytics, machine learning, and innovation, building internal tools and dashboards for energy-market and R&D use cases.

    • Python
    • SQL
    • Power BI
    • AWS
    • Git
    • Machine Learning

Blog

Notes on data engineering, building products, and what I learn along the way.

First posts coming soon.

Contact

Let's build something.

Whether it's a data role, a product idea, or a hackathon team, my inbox is open. Email is the fastest way to reach me.