Software engineer. From Melilla, Spain.
I write full-stack code, lift things, and occasionally talk to people.
CS graduate from the Universidad de Almería (Software Engineering track). I build backends with real domain complexity and ship them end to end — schema, API, auth, tests, pipeline, deployment. Not portfolio projects nobody uses: software with real users and real constraints. I care about clean architecture, testable code, and not over-engineering stuff that should be simple.
In practice, that's looked like: a competition platform tested with 33 real participants across 5 user roles · CI/CD with Playwright e2e and GitHub Actions · a data pipeline pulling Wikidata + Wikipedia through Redis/BullMQ job queues · owning team workflow and deployment on two group projects.
Competition management platform for amateur football-7 leagues in Melilla. Born from the frustration of watching real leagues run on WhatsApp groups and spreadsheets. Built as my final degree project and taken through a full user testing round — 33 participants across five roles, with the survey results feeding back into the product.
Stack: NestJS · Prisma · PostgreSQL · React · Docker · Cloudflare R2
Not a tutorial project. Role-based guards and ownership patterns, TDD, injectable time dependencies, object storage, Playwright e2e tests, CI/CD, and the usual production headaches.
The live test deployment is now closed. The full set of artefacts — final system screenshots, sketches, prototypes, UML diagrams and documentation — is published at mtzgutierrez.github.io/TFG-MartinezGutierrezRaul.
- I ship with users watching. Gangs League went through a structured testing round with 33 real participants; the feedback turned into concrete product changes, not a slide.
- I take the unglamorous half. On Versus and NRP Studio I owned the team workflow, the CI/CD pipelines and the deployment — the part that decides whether the rest of the team can move.
- I document so someone else can pick it up. UML, ER diagrams, prototypes and written specs, published rather than kept in a folder.
- I explain trade-offs, not just decisions. Layered architecture, TDD, injectable dependencies — each because of a specific problem, and I can tell you which one.
- I finish things. Every project here reached a working, deployed state.
Things I know well enough to have strong opinions about:
Things I also use:
🆚 Versus — team work & infrastructure
Trivia game where you don't just answer — you bet. Real-time duels with several game modes, fed by real-world numbers scraped off the web so the questions never repeat. Team project: I owned the team workflows, the automations (CI/CD pipelines), and the deployment. The server isn't online at the moment, but the work behind it stands.
Stack: Spring Boot · Angular · PostgreSQL · Scrapy · Docker
⚔️ Bellum Atlas — data pipelines & external integrations
Interactive historical atlas of battles on a map. It pulls open data from the Wikidata and Wikipedia APIs, places each battle on a full-screen map with its own rich profile, and layers AI-generated narratives on top — including daily entries — to turn scattered history into something you can explore. Ingestion runs on Redis/BullMQ queues. The deployment is down, but the code and the pipeline are in the repo.
Stack: NestJS · React · Prisma · PostgreSQL · Redis (BullMQ) · Leaflet · Docker
📋 NRP Studio — fullstack CRUD & automation
Software requirements management tool, a smaller team project. Hands-on fullstack work (projects, clients, solutions and requirements with JWT auth) plus the automation side to keep the pipeline moving.
Stack: Spring Boot · React · MySQL · Docker
Degree finished. Looking for work on backends with real domain complexity, data integration and ingestion pipelines, or products that already have users and need someone to make them hold up.
Contact: martinezgutierrezraul@outlook.es · LinkedIn



