ibai.dev
Open to backend / AI roles · Remote-friendly

Ibai Mutiloa Aliaga.

Backend Engineer · AI Systems · Cloud Infrastructure

I build production-grade backend systems and AI pipelines — from RAG architectures with semantic search to cloud-native infrastructure with proper observability.

View ProjectsGet in touchGitHub ↗
BackendPythonFastAPINode.jsPHPREST APIsPostgreSQL
AI / MLRAG SystemsLLM IntegrationpgvectorSemantic SearchScikit-learn
InfraDockerAzureLinux / DebianJenkinsOIDC / MFAGrafana
// Selected Projects
AI InfrastructureProduction

University Intranet RAG System

End-to-end Retrieval-Augmented Generation system deployed on university intranet. Document ingestion pipeline, vector storage with pgvector, semantic retrieval, and LLM response synthesis with context grounding.

200+Daily users
pgvectorVector store
FastAPIPostgreSQLpgvectorDockerAzureLLM
↗
Security / AuthProduction

Modern MFA & OIDC Implementation

Designed and implemented multi-factor authentication and OpenID Connect single-sign-on for institutional users. Replaced legacy session auth with token-based flows and identity federation.

OIDCAuth protocol
TOTPMFA method
0Incidents post-deploy
OIDCTOTPJWTLinux
↗
ML / PredictionSide Project

LEZGuard — Emissions Prediction

Machine learning pipeline to predict vehicle emissions and compliance with Low Emission Zone regulations. Feature engineering from traffic and environmental data, model evaluation and API exposure.

MLPipeline
RESTAPI layer
CI/CDJenkins
PythonScikit-learnFastAPIJenkinsDocker
↗
Full Stack PlatformSide Project

Solraise — Web Platform

Full-stack web platform with a modern backend architecture, user management, observability stack (Grafana + Matomo), and containerised deployment. Designed for horizontal scaling from day one.

Node.jsBackend
DockerContainers
Node.jsDockerGrafanaMatomoPostgreSQL
↗
LMS / Plugin DevProduction

Moodle Plugin Development

Custom Moodle plugins extending LMS functionality for institutional workflows. PHP backend, Moodle APIs, database abstraction layer and integration with external university systems.

PHPBackend
MoodlePlatform
APIsExternal integration
PHPMoodle APIPostgreSQLREST
↗
DevOps / ObservabilityProduction

Observability Stack

Designed and deployed metrics and analytics infrastructure for production services. Real-time dashboards, alerting rules, and user analytics for data-driven decisions on system health.

Real-timeMonitoring
AlertsThreshold rules
SLODefined & tracked
GrafanaPrometheusMatomoDockerLinuxJenkins
↗
// Experience
sept. 2025 — Present
University / Institutional
Software Developer | Azure, Docker, RAG Systems

Development of a Retrieval-Augmented Generation (RAG) system integrating university regulations into a legacy intranet used daily by 200+ employees. Implemented vector search with pgvector, secured platform access with OIDC/MFA, and set up observability and analytics using Grafana and Matomo. Collaborated with cross-functional teams and supported CI/CD and containerized deployments.

PythonpgvectorRAGAzureDockerGrafanaMatomoOIDCPostgreSQL
oct. 2024 — July 2025
University / Institutional
Software Development Intern | PHP, Moodle, Grafana

Software development intern contributing to web applications and Moodle-related features. Supported PHP-based developments, improved functionality and user experience for students and staff, and assisted with maintenance and deployments.

PHPMoodleGrafanaDockerPostgreSQL
// About
// Background

I'm a software developer based in Spain, focused on the engineering side of things: backend systems, AI integrations, and the infrastructure that keeps them running. I studied Computer Engineering and have been building real production systems since early in my career.

My work sits at the intersection of solid backend engineering and applied AI — building the pipelines, APIs, and databases that make LLMs actually useful in production, not just in demos.

// Engineering principles
→Observability is not optional. Instrument everything, alert on what matters.
→Boring technology beats clever technology. Use Postgres, Docker, and proven stacks before exotic ones.
→RAG is an architecture problem, not a prompt problem. Retrieval quality determines output quality.
→Security is a feature. MFA, least-privilege, and proper token handling from the start.
→Latency budgets are real constraints. Design APIs around the P95, not the average.
// Currently exploring
→Agentic AI systems and tool-use patterns with LLMs
→Advanced vector search tuning and hybrid BM25 + semantic retrieval
→Kubernetes for container orchestration beyond basic Docker Compose
→Event-driven architectures with message queues
// Looking for

Roles where I can go deep on backend systems, AI/LLM infrastructure, or cloud-native architecture. I thrive in environments that value technical depth, good engineering practices, and building things that run reliably at scale.

Open to remote, hybrid, or on-site positions in Spain or internationally. Comfortable in English and Spanish.

// Contact

Let's talk engineering.

Whether you have a backend or AI role in mind, want to discuss architecture, or just want to chat about RAG systems and vector search — feel free to reach out.