Billing & Payments APIs
Designed mission-critical mobility billing APIs with event-driven architecture, operational observability, and cloud-native delivery.
JavaSpring BootKafkaAWSKubernetes
Open to senior roles
Diego Barrio H. Senior Backend Engineer Java, Spring & Applied AI
15+ years building scalable, maintainable, and observable backend platforms for banking, fintech, and mobility. I now bring that same engineering standard to RAG pipelines, LLM integration, and agentic workflows.
15+
years shipping production software
4
critical industries: banking, fintech, mobility, enterprise
3
AI products built end to end, from RAG pipeline to deploy
Track record at Europcar Mobility GroupAccentureBanco SabadellSABISCaja Mediterráneo
01
Experience in systems that cannot fail.
Experience 2024 — Present ● Current
Senior Backend Engineer
Mindden / Europcar Mobility Group
Built business-critical billing and payments APIs, strengthening operational robustness and delivery velocity.
JavaSpring BootKafkaAWSKubernetes
2022 — 2024
Senior Analyst
Accenture
Delivered microservices and data optimization for banking with focus on performance, uptime, and maintainability.
JavaSpringMicroservicesOracle
2014 — 2022
Senior Specialist
SABIS
Designed enterprise platforms and internal integrations to standardize workflows across teams.
Spring MVCJPASharePoint.NET
2007 — 2014
Analyst / Developer
Banco Sabadell · Caja Mediterráneo
Evolved high-criticality online banking and ATM transactional platforms.
JavaOnline bankingATM
02
Real work in complex domains.
ProjectsCore Banking Modernization
Migrated core banking processes to modular services, improving performance, traceability, and time-to-market.
SpringHibernateOracleSQL ServerCI/CD
Enterprise Platforms
Built internal platforms and integrations that reduced operational friction and technical debt.
Spring MVCJPASharePoint.NETOracle
03
Technologies and practices to scale both product and team.
StackBackend Core
JavaSpring BootSpring MVCSpring SecurityJPA/HibernateREST APIs
Architecture & Scale
MicroservicesDDDHexagonalEvent-drivenClean ArchitectureSRE mindset
Cloud & Platform
AWSKubernetesDockerApache KafkaCI/CDGitLab/Jenkins
Data & Observability
PostgreSQLOracleSQL ServerMongoDBGrafanaSplunk/Sentry
Applied AI
LLM integrationRAGEmbeddingsAgentic workflowsTool callingMCPOpenRouterOllama
04
Where I put AI to work.
LabPersonal products I design, build, and run myself. They are pre-production and I am their first user, which is exactly why they are where I test architecture decisions before they reach a client system.
BACKEND To the future Akademia — RAG-generated exams from your own material
The hard part is the RAG generation itself: turning the resources a user uploads into tests and questions automatically. Built on the OpenAI embeddings API through OpenRouter, with Gemini Embedding 2 support in progress.
JavaSpring BootPostgreSQLRAGOpenRouterStripe
TokenMeter — What code generation actually costs
An open exploration of tokenization and LLM cost in code generation, across both agentic loops and direct single-shot flows. Multi-provider pricing comparison with dynamic price synchronization.
JavaSpring BootDockerCloudflareLLM APIs
Forma — A validated contract for LLM-generated plans
A personal fitness operating system: body composition, training, nutrition and shopping cost in one weekly loop. The interesting part is the boundary I drew around the model. An LLM can generate a nutrition plan, but only from a catalog the API publishes, and any food it invents is rejected on import. Macros are measured data, not something a model gets to estimate.
Java 21Spring Boot 3HexagonalPostgreSQLFlywayReact 19Playwright
Local AI Lab — Self-hosted models on an RTX 3060
Local LLM deployment on 12 GB of VRAM with Ollama and LM Studio. Where I evaluate Qwen, Gemma, Llama, DeepSeek R1 and Mistral Nemo against agentic development workflows and MCP integrations.
OllamaLM StudioMCPClaude Code
05
Let's build something that cannot fail.
ContactAvailable for senior backend roles, remote across the EU, and for consulting on applied AI in existing systems. I reply within 24 hours.