Applied AI Engineer, Tarragona
hace 20 días
Tarragona
Applied AI Engineer Applied AI Engineer Country: Spain IT STARTS HERE Santander (www.santander.com) is evolving from a global, high-impact brand into a technology-driven organization, and our people are at the heart of this journey. Together, we are driving a customer-centric transformation that values bold thinking, innovation, and the courage to challenge whats possible. This is more than a strategic shift. Its a chance for driven professionals to grow, learn, and make a real difference. Our mission is to contribute to help more people and businesses prosper. We embrace a strong risk culture and all our professionals at all levels are expected to take a proactive and responsible approach toward risk management. Our Chief Data Artificial Intelligence Officer (CDAIO) division is building a world-class AI Data team to make a difference in the lives of over 170 million people worldwide, through one of the largest banks in the world. We are undergoing one of the biggest transformations in our history and technology is at the heart of our strategy. Join our team to play a part in one of the most important technological projects for the financial sector in the world. THE DIFFERENCE YOU MAKE Santander AI Lab (CDAIO) is looking for an Applied AI Engineer based out of Madrid, Spain. The AI Lab is the applied innovation engine of one of the worlds largest banks. We detect emerging opportunities, build working prototypes, validate them with real data, and transfer them to scale. We work with Anthropic, Sakana AI, AWS, ICMAT, CMU, INRIA and other world-class partners. Our published research (arXiv:2602.14606, arXiv:2602.02170) sets the formal foundation for everything we build. Were shaping the way we work through innovation, cutting-edge technology, collaboration and the freedom to explore new ideas. To succeed in this role, you will be responsible for: Designing, building and deploying production-grade agentic AI systems - multi-agent orchestration with real memory, planning, tool use and error recovery. Not demos. Systems that work. Developing and fine-tuning small and medium language models (SLMs) for regulated banking use cases, including custom evaluation frameworks and domain-specific benchmarks. Architecting and implementing MCP servers, A2A protocols, and federated API layers that allow AI agents to operate across the groups multi-country infrastructure. Prototyping new ideas in two-week sprints: hypothesis, architecture, code, functional demo, one-pager. This is the labs operating rhythm - you need to thrive in it. Collaborating with researchers, data scientists and business stakeholders to translate complex technical concepts into tangible bank value - Alchemy-style transformations of legacy assets. Keeping the lab at the frontier: monitoring emerging research, evaluating new tools (Harness Engineering, Kiro, Windsurf, Devin), and integrating them into the labs workflow when they add real value. Producing clean, tested, observable code that can be handed off to the AI Science team for production scaling. You own your code end-to-end. WHAT YOULL BRING Our people are our greatest strength. Every individual contributes unique perspectives that make us stronger as a team and as an organization. Were enabling teams to go beyond by valuing who they are and empowering what they bring. The following requirements represent the knowledge, skills, and abilities essential for success in this role. Reasonable accommodations may be made to enable individuals with disabilities to perform the essential functions. Professional Experience 4-8 years of software engineering or AI engineering experience, with at least 2 years building and maintaining LLM-powered systems in production environments - not in notebooks. (Required) Demonstrated hands-on experience designing and deploying multi-agent AI systems with real-world complexity: memory management, stateful orchestration, tool use, multi-step planning, and graceful failure recovery. (Required) Experience fine-tuning or adapting language models (SFT, LoRA, RLHF) for domain-specific tasks, including dataset curation and evaluation design. (Required) Track record of delivering complete systems independently within tight timelines - from architecture decision to production-ready code. Portfolio of real systems, not just demos. (Required) Experience building and consuming REST APIs and integrating AI systems with enterprise data sources, cloud services and third-party platforms. (Required) Prior experience in banking, fintech, or other regulated industries. (Preferred) Exposure to Harness Engineering methodologies: spec-driven development, AI-assisted software creation at scale. (Preferred) Education Bachelors or Masters degree in Computer Science, Software Engineering, Mathematics, Physics, or equivalent technical field. (Required) Masters degree or equivalent advanced qualification in AI, Machine Learning or related discipline. (Preferred) Languages English: professional working proficiency - all technical documentation, papers and partner communications are in English. (Required) Spanish: professional working proficiency - day-to-day team communication and stakeholder collaboration. (Required) Hard Skills Python: advanced proficiency. Clean, tested, production-grade code. You know the difference between a script and a system. (Required) Agentic AI frameworks: LangGraph, AutoGen, CrewAI, or Semantic Kernel. You have designed multi-agent workflows with these tools, not just run their tutorials. (Required) LLM APIs and model ecosystems: Anthropic Claude API, OpenAI, open-source models (Llama, Mistral). You understand cost, latency and quality trade-offs. (Required) Cloud infrastructure: AWS (Bedrock, Lambda, SageMaker, S3). Comfortable deploying and monitoring AI systems in cloud environments. (Required) Model evaluation and observability: you design evals, not just run them. Experience with LLM-as-judge, RAGAS, Promptfoo, Langfuse or equivalent. (Required) API development: FastAPI or equivalent. You can build a production-ready API around an AI system. (Required) DevOps basics: Docker, Git, CI/CD pipelines. Your code ships, not just runs locally. (Required) MCP (Model Context Protocol) server design and implementation. (Preferred) SLM fine-tuning pipelines: vLLM, Ollama, Unsloth or equivalent for efficient domain adaptation. (Preferred) Knowledge graphs or GraphRAG for structured knowledge retrieval in complex domains. (Preferred) A2A protocol, x402 or AP2 for agentic payments or agent-to-agent communication. (Preferred) Kubernetes, Terraform or equivalent for production-scale deployment. (Preferred) Soft Skills Radical autonomy: you arrive on Monday with your own priorities. You dont wait to be told what to build - you propose it. Builders mindset: when you encounter an interesting idea, your first instinct is to build a proof of concept, not write a slide about it. Speed with judgment: you can deliver a functional demo in two weeks and know when a prototype is ready to transfer versus when it needs more work. Frontier awareness: you read papers the week they drop. You have opinions about what Anthropic, Sakana AI and AWS are building. You learn by doing, not by watching courses. Communication across roles: you can explain a complex agentic architecture to a business stakeholder and write a technical one-pager for the banks leadership team. Both matter here. Collaborative rigor: you give and receive direct technical feedback. You document your decisions not because someone told you to, but because future you - and your teammates - will need it. Python, LangGraph, AutoGen, Anthropic, Sakana, LLM
