Alberto Guinda Sevilla
AI Engineer · LLM Agents · Computer Vision · Digital Twin
I build software with artificial intelligence inside it, and I leave it running in production — not in a demo. I work end to end on my own: I talk to whoever has the problem, design the solution, write it and maintain it. Where I am hardest to replace: 3D replicas of buildings and infrastructure (digital twins), cameras that understand what they see, and sensors that warn you before something fails.
🏗️ WHAT I DO RIGHT NOW: at OFITECO (VINCI Group) I am the only developer on the platform that watches over critical civil works. A 3D viewer of the project in the browser that used to take 15 seconds to open and now takes 1, fifty solar-powered cameras that spot on their own what should not be there, and a structural-health system already installed at the client site.
🤖 MY OWN PRODUCTS, SHIPPED: outside work I design, build and publish my own tools. Five are live on the internet today, with a domain, a real payment gateway and users: a home-buying copilot driven by an AI agent, a booking system that runs over WhatsApp, a CV search engine that sends no data to any cloud, a health app for elderly users, and a suite of 20 developer utilities.
📊 MEASURABLE RESULTS: at Pikolín I cut sales-results query time by 40% (I have the letter from the CIO to show for it). At UB Manufacturing I resolved more than 200 technical issues a month. At OFITECO, a viewer that was unusable now opens fifteen times faster.
EDUCATION
Master's in Digital Transformation and the Internet of Things — Instituto Tecnológico de Aragón
Grade average 8.6/10. Expert in AI Software Development from BIG School and Universidad Internacional Isabel I de Castilla. Higher Diploma in Web Application Development. 12 Anthropic Academy certifications and 18 Google Cloud Skills Boost badges.
See every certification and degreeTECH STACK
Technologies I work with
DIGITAL TWIN · BIM · GIS
AI · ML · COMPUTER VISION
IOT · HARDWARE
FRONTEND · 3D
BACKEND
DATABASES
DEVOPS · QA
What I am on right now
AI agents that carry out real tasks and ask permission before charging anyone, built with LangGraph and full traceability. Stripe payments reconciled against webhooks. Edge deployment on Cloudflare Workers. And photorealistic Google 3D maps you can walk through, right into a flat.
Letters of recommendation
References from ITA and Pikolín, available right here
View references (opens in a new tab)Professional work and personal builds — from applied AI and computer vision to digital twins and industrial IoT.
Live right now
These are not mockups or videos: they are websites and apps that are running, with their own domain and real users. Open them and see for yourself.
A LangGraph agent with 19 tools and human confirmation before any payment, over a transactional backend with hard guarantees and a 3D viewer you can walk through.
Digital Twin · BIM/GISWeb-based digital-twin platform (BIM + GIS + IoT) for monitoring critical infrastructure in real time. Built solo, from architecture to production.
Web platform that auscultates bridges and buildings via vibrations: automatic operational modal analysis (OMA), tracking and damage alerts. In production.
Computer Vision · IoTFull-stack service to operate autonomous 4G cameras (solar, no fixed IP) and analyse events with computer vision over distributed critical infrastructure.
Modern-stack rewrite of OFITECO's infrastructure auscultation app: field readings, calibrations, issues, alarms and SCADA, offline-first, for Android and iOS.
Strategic-management panel for infrastructure monitoring: advanced data table, cascading filters, reactive KPIs and interactive maps, in 4 languages, WCAG AA.
Mobile · React NativeAndroid IPTV app with 226 free DTT channels and on-device AI recommendations (no cloud). 79 languages, clean architecture, native Kotlin modules.
Whether you're hiring or have a project in mind, drop me a line.
Let's talkI build, single-handedly and from architecture to production, OFITECO’s (VINCI Group) digital-twin platforms for monitoring critical civil infrastructure (dams, tunnels, bridges).
xdata-bim digital twin (BIM + GIS + IoT in the browser): IFC viewer with web-ifc (WASM) and a .frag binary format (load time ~15 s → ~1 s), real-time monitoring over WebSocket and anomaly detection (Z-score, Bollinger, regression, correlation). 37 components, ~840 i18n keys (4 languages), E2E with axe-core (WCAG AA).Tech stack
UB Manufacturing S.L.
Analysis and assessment of technology infrastructure, focused on modernising and digitalising business systems.
Tech stack
Building advanced IoT solutions, digital twins and computer-vision web applications for research and innovation projects.
Tech stack
Agencia Digital SP
Building custom web solutions for clients, focused on WordPress and the integration of AI technologies.
Tech stack
Building internal tools for business data analysis and forecasting, focused on process optimisation and performance.
Tech stack
Tickamore (Sicomoro)
Specialised technical support for ticketing solutions, focused on incident resolution and advanced database management.
Tech stack
Two signed reference letters from the people who managed me. Read or download the originals.
“Destacó por su notable proactividad, buscando siempre ir más allá de lo esperado y anticipándose a las necesidades de los proyectos. […] No dudo en recomendarlo sin reservas.”
Adán Piñero
Multi Project Manager, Cognitive Systems & Big Data — ITA
“Alberto destaca por cualidades como dedicación, constancia, organización y capacidad de trabajo. […] Muy buen candidato para ocupar cualquier puesto relacionado con sus capacidades.”
José Ignacio Brieba
CIO, Grupo Pikolin
What I learn building applied AI, RAG, computer vision and digital twins — explained plainly, with production-grade judgement.
Generative engine optimization (GEO): llms.txt, schema.org (Person, FAQPage, hasCredential), extractable content and E-E-A-T so AI cites your website.
How to build a real-time IoT dashboard: MQTT/P2P ingestion, time-series storage, live visualization, smart alerts, and turning raw data into a useful decision.
How I built a free LLM fallback chain (Groq, OpenRouter, Pollinations) with graceful degradation: chained free tiers that survive quota limits and timeouts.
I've trained this assistant on my CV and portfolio. Ask it about my experience, my stack — or why I'd be a great fit for your team.
What recruiters and clients usually ask me — answered plainly.
Alberto Guinda Sevilla is an AI Engineer and full-stack developer based in Zaragoza (Spain). He takes AI systems to production —computer vision, LLM agents and digital twins— from architecture to deployment. He currently works at OFITECO (VINCI Group), building single-handedly over critical infrastructure.
His niche is digital twins over BIM and GIS, computer vision at the edge (YOLOv8, CLIP), LLM agents with Claude Code, and industrial IoT, applied to monitoring critical civil infrastructure.
He is based in Zaragoza, Spain. He works remotely and is available for on-site or hybrid projects in Zaragoza, Madrid and Barcelona.
Yes. He is open to full-time roles, collaborations and freelance projects, both remote and on-site. His CV (ES/EN) is downloadable on the site and he replies within 24 hours.
Day to day it is TypeScript, React 19, Svelte 5 and Vite on the front end, and Python with FastAPI on the server. Around that: Astro and Next.js for the web, Three.js and CesiumJS for 3D, PostgreSQL, Docker and deployment on Cloudflare. On the AI side, agents with LangGraph and Claude Code, RAG with embeddings, and computer vision with YOLOv8 and CLIP. On mobile, React Native and Capacitor.
Every project starts from his own scaffold with quality already wired in: layered architecture, strict typing with no shortcuts, automated tests, design tokens with contrast verified against WCAG, and a gate that blocks the push if anything breaks. That means zero type errors, zero dependency vulnerabilities and verified accessibility before delivery, not after. It is not a promise — they are checks that run themselves on every change.
A Master's in Digital Transformation and the Internet of Things from Instituto Tecnológico de Aragón (8.6/10), an Expert degree in AI Software Development from BIG School and Universidad Internacional Isabel I de Castilla, and a Higher Diploma in Web Application Development. On top of that, 12 Anthropic Academy certifications (Claude Code, MCP, agents, AI Fluency) and 18 Google Cloud Skills Boost badges, with the Associate Cloud Engineer exam in preparation.
I'm open to freelance projects, collaborations and full-time roles.
I reply within 24 hours