Now · AI Architect & Engineering Lead at Aareon

I build AI agents that ship to production. 

AI Architect & Engineering Lead at Aareon. I design agentic systems and lead the teams that take them from architecture to production at enterprise scale.

Portrait of Miguel Abdón

01Selected work

Things I've built.

Enterprise agent platforms at work, and open-source experiments on GitHub.

01Santander · Enterprise

Multi-agent Payment Platform

A production multi-agent platform where autonomous agents collaborate across several Santander brands to run end-to-end payment workflows. Built as a reusable global capability so business teams can deploy new payment agents quickly.

  • Multi-agent
  • Payments
  • Enterprise scale

Enterprise work, source is private

Fig. 01 — Brand agents, one orchestrator

02Open source · WebSockets

Real-time Customer Support (source on GitHub, opens in a new tab)

Support chat run by LangGraph agents over WebSockets, with a Next.js front end. Modular, scalable and built to automate real support workflows.

  • LangGraph
  • TypeScript
  • Next.js
  • PostgreSQL
Fig. 02 — Router and specialist agents

03Open source · Retrieval

Dynamic Content RAG (source on GitHub, opens in a new tab)

A backend that writes product copy, social posts and summaries grounded in live data from multiple sources.

  • FastAPI
  • PostgreSQL
  • Redis
  • Docker
Fig. 03 — Retrieve, then generate

04Open source · Gen AI

Resume Analyzer (source on GitHub, opens in a new tab)

Gemini scores resumes against job descriptions through a node-based pipeline, processed in parallel on Kubernetes, with a chat to explore the results.

  • FastAPI
  • React
  • Pydantic AI
  • Kubernetes
Fig. 04 — Parallel scoring pipeline

02About

From architecture diagram to production, with the team that gets it there.

I specialise in enterprise AI, agentic systems and modern software engineering. I lead multidisciplinary teams while keeping deep hands-on ownership of architecture, development, DevOps, deployment and observability.

My focus is the gap between a promising prototype and a reliable system: bridging technical execution, engineering leadership and business requirements until it runs in production.

Based in
Madrid, Spain
Studied
Software Engineering, U-tad
Speaks
Spanish · English C1 · French B1
Fig. 05 — The stack, top to bottom
  1. 06LeadershipEngineering leadership · Mentoring · Technical direction
  2. 05Agents & LLMsLangGraph · LangChain · MCP · A2A · RAG · Agentic coding
  3. 04Machine learningPyTorch · TensorFlow
  4. 03LanguagesPython · TypeScript
  5. 02DataPostgreSQL · MongoDB · Redis · Neo4j
  6. 01PlatformAWS · Azure · Docker · Kubernetes · CI/CD · Observability

03Experience

Where I've worked.

  1. Nov 2025 — Present

    Current role

    AI Architect & Engineering Lead

    Aareon Group

    Lead AI architecture and a large multidisciplinary engineering team, combining hands-on technical depth with people leadership.

    • Own the end-to-end lifecycle of enterprise AI applications: system design, development, CI/CD, cloud infrastructure, deployment, observability and production reliability.
    • Design production-grade agentic architectures that integrate LLMs, autonomous agents, MCP, enterprise systems, APIs and data platforms.
    • Drive adoption of AI-native engineering practices, from modern frameworks to AI coding systems and agentic development workflows.
    • Mentor engineers, set technical direction and translate business strategy into architecture and delivery decisions.
  2. Jul 2024 — Nov 2025

    AI Labs Software Engineer

    Santander Digital Services

    Co-founded AI Labs to build impactful generative AI and presented solutions to senior leadership.

    • Designed and delivered a production multi-agent payment platform used across multiple Santander brands.
    • Designed a custom LangChain SDK for multi-agent automation on AWS and Azure, building MCPs and A2A protocols to orchestrate large-scale document workflows.
    • Collaborated with IBM on AI for legal and regulatory text analysis.
    • Ran discovery workshops with business, product and leadership to turn business processes into production AI agents.
  3. Sep 2023 — Jul 2024

    IT Engineer

    KPMG Spain

    Backend systems with a focus on database architecture and data processing pipelines.

    • Supported deployment and performance tuning of client-facing applications in production.

04Contact

Let's build something.

Have a project, an idea or just want to talk about AI agents? Drop me a line.