ÁlvaroSuárezRésumé
Menu

Álvaro Suárez, Madrid

I turn live data into systems people can depend on.

Senior Software & Data Engineer with eight years of experience building backend platforms, real-time pipelines and applied AI products.

Explore selected workStart a conversation
Portrait of Álvaro Suárez outdoors

Senior Software & Data EngineerMadrid, Spain

A data path from live sources to useful productsData enters from APIs and events, is normalized and processed, then reaches products through data models and services.SourcesProcessModelServeBATCHSTREAMEXPLAIN

8 yearsfrom Java services to cloud data platforms

~100MAPI requests served each month

~4 GB → ~8 min60 fps CSV converted to Sony-compatible MPEG-DASH

Selected systems

Engineering work measured in movement.

Data arriving, changing shape and reaching the people or products that need it. These are the systems behind that path.

LaLiga Data Entities / Sportian Data Hub

Unified sports master data

2022-2025

An organization-wide master data platform that brings together sports data from multiple providers, normalizes it and assigns canonical entity IDs so every product can work from the same model.

Provider dataNormalizeCanonical IDsOrg-wide APIs

My contribution

I designed and built Python microservices for push and pull ingestion, entity resolution and API delivery. Provider-specific identifiers were reconciled against canonical IDs, allowing projects across the organization to consume different data origins transparently through one contract.

  • Python
  • AKS
  • Event Hub
  • SQL Server
  • Cosmos DB
  • REST

Tracab / Sony Hawk-Eye integration

60 fps tracking, delivered in minutes

2025-2026

A compatibility layer that made a new skeletal tracking feed usable by an existing post-match visualization system without changing the consumer.

Tracab CSVStreamMPEG-DASHHawk-Eye

My contribution

I worked on the streaming microservices pipeline that ingested frame-level CSV data for players, officials and the ball, then converted it into the MPEG-DASH contract expected by Sony Motion Server.

  • Python
  • FastAPI
  • AKS
  • Azure Storage
  • MPEG-DASH

LaLiga Technical Referees Committee

Explainable referee assignment

2026

An AI-assisted system that produces three viable officiating-team proposals for every first and second division matchday.

Referee dataConstraintsOptimizeExplain

My contribution

I built a scheduled Python workflow that scores assignments from performance history, recent appointments, physical status and conflict constraints. An LLM turns the underlying evidence into a reviewable explanation for each proposal.

  • Python
  • Kubernetes
  • Azure SQL
  • FastAPI
  • LLM

PEAK Sports and Media

A data platform for people and agents

2026-present

A Databricks platform that turns multiple data sources into governed analytical datasets and a shared context for AI agents.

SourcesBronze / silverGoldBI + agents

My contribution

I develop ingestion and transformation pipelines across a medallion architecture with Python and PySpark. Curated gold tables power Tableau dashboards and agents that answer questions across the platform.

  • Databricks
  • PySpark
  • Python
  • Tableau
  • AI agents

ISEO / Independent work

Energy data from the ground up

2024-2025

A cloud data platform built from scratch around the Spanish energy system for an early-stage company.

Energy dataAzureModelPower BI

My contribution

As a freelance software engineer and cloud architect, I designed the Azure foundation, Python data workflows and the path from collected data to Power BI reporting.

  • Azure
  • Python
  • Data pipelines
  • Power BI

Experience

Built from the backend out.

A path from Java services and financial applications to leading backend teams and building data-intensive products in the cloud.

  1. Oct 2022-present

    Senior Software Engineer

    Globant / Sportian

    Backend and data systems for sports products: master data, live tracking, optimization and Databricks platforms on Azure.

  2. Jan 2024-2025

    Senior Software Engineer & Cloud Architect

    Independent

    Cloud and data architecture for ISEO's Spanish energy data platform, developed alongside my main role.

  3. Dec 2020-Oct 2022

    Backend Lead

    KANARA Sportech

    Built the company's backend and wearable-data ETL on Google Cloud, then led a team of four backend engineers.

  4. Aug-Dec 2020

    Senior Software Engineer

    Independent

    Delivered a synchronization service between Google Sheets and PostgreSQL for a telecommunications client.

  5. Mar-Sep 2020

    Software Engineer

    Quirumed

    Developed e-commerce services for product translation and stock control.

  6. Apr 2019-Mar 2020

    Junior Software Engineer

    Almis

    Developed web applications for financial analysis.

  7. Sep 2018-Apr 2019

    Junior Software Engineer

    El Corte Inglés Seguros

    Developed Java 8 and Spring Boot backend services.

Capabilities

The stack follows the problem.

My core is Python, Java and cloud data engineering. The useful part is knowing where each tool belongs in a working system.

Backend systems

APIs, microservices and asynchronous workloads designed around explicit contracts and observable failure modes.

Python, Java, JavaScript, FastAPI, Flask, Spring Boot, Asyncio, Celery

Data platforms

Streaming and batch pipelines that move from raw events to trusted datasets and products.

PySpark, Databricks, Pandas, Polars, NumPy, medallion architecture

Cloud delivery

Containerized systems running on managed cloud infrastructure, with practical networking knowledge underneath.

Azure, Google Cloud, Docker, Kubernetes, AKS, Linux, HTTP

Data movement

Event-driven integration across providers, services and storage with real-time and near-real-time requirements.

Event Hub, Kafka, RabbitMQ, Google Pub/Sub, REST, requests, httpx

Persistence

Relational and document models selected for access patterns rather than fashion.

PostgreSQL, MySQL, Azure SQL, SQL Server, MongoDB, Cosmos DB

Applied AI

Agentic workflows, grounded explanations and local model experimentation with engineering guardrails.

AI agents, LLM integration, llama.cpp, LM Studio, GGUF, Llama, Mistral

In the open

Research made reproducible.

My undergraduate thesis is developed in public, alongside focused Python tools built to solve real engineering friction.

View GitHub profile
Featured · Undergraduate thesis

Generating and detecting fallacies with LLMs

A reproducible, resumable Python pipeline that asks local and hosted LLMs to generate logical fallacies, then evaluates every output across configurable judge models. Versioned settings and JSONL artifacts keep experiments comparable and auditable.

Generate → Evaluate → ComparePython package

async-watchdog

A dependency-free asyncio heartbeat monitor with sync and async timeout callbacks, restartable lifecycle and defensive logging.

macOS CLI

piximport

A Python tool that detects SD cards and organizes photos by EXIF date, camera make and media type.

Data project

SpainRentalHeatMaps

A Flask application that collected rental listings through the Idealista API and rendered geographic heatmaps.

About

I care about what happens between a clean diagram and production.

I am a software and data engineer based in Madrid. I enjoy the less visible engineering that makes products trustworthy: explicit contracts, reliable data movement, useful observability and systems that remain understandable as they grow.

Outside delivery work, I experiment with local open-weight models, agentic development workflows, networking, security and OSINT. Photography is the non-software practice I keep returning to; it is also why tools like piximport exist.

Education

Bachelor's Degree in Computer Engineering

Universidad Politécnica de Madrid · 2014-2020

Undergraduate thesis: a framework for evaluating how well different LLMs generate and detect argumentative fallacies.

Explore the public thesis project

SpanishNative

EnglishProfessional working proficiency · B2-C1

Contact

Have a hard data problem?

[email protected]

Madrid, Spain · Open to engineering conversations worldwide.