Forward Deployed AI Engineer

Job description

¿Cómo será tu día a día en Tuio? You will work directly with business teams to understand the processes, decisions, systems, data, exceptions, and risks involved in each use case. You will design the architecture of agent-based systems, selecting the most appropriate models, patterns, and components for each problem. You will build applications powered by language models using context engineering, tool calling, structured outputs, RAG, and multi-step workflows. You will experiment with models, instructions, and architectures to optimize the balance between quality, latency, cost, and reliability. You will connect agents to internal and external tools through APIs, MCP servers, databases, and SaaS platforms. You will rapidly create prototypes and evolve them into solutions used with real-world data, systems, and use cases. You will design evaluations, test cases, and regression tests, and instrument agents to monitor their traces, errors, latency, and costs. You will implement reliability and control mechanisms such as permissions, validations, retries, fallbacks, action limits, and escalation to a human. You will deploy and operate solutions using testing, version control, CI/CD, and sound engineering practices. You will turn the lessons learned from each deployment into reusable components, evaluations, and patterns for future Tuio agents. You will become embedded in one of Tuio’s priority domains to understand its processes, systems, and key technical challenges. You will take responsibility for one or two use cases and bring at least one of them from prototype to a pilot with real users, defining its evaluations, observability, and next steps for scaling. Your day-to-day work will combine close collaboration with the business, end-to-end technical ownership, and a way of building based on experimenting quickly, measuring rigorously, and strengthening controls whenever the level of risk requires it. You will be involved from the discovery stage and work directly with the people who understand and operate each process. Depending on the project, you will collaborate with AI Managers and colleagues from Business, AI, Data, and Engineering to define and evolve the solution. You will take end-to-end technical ownership, including AI architecture, code, integrations, evaluations, deployment, observability, and reliability. In some projects, you will also play a significant role in discovery and functional design. You will build small initial versions to validate hypotheses with real users, measure outcomes, and iterate quickly. We will consciously accept technical debt when it helps us learn, but we will strengthen the architecture, testing, and security before scaling. You will treat agents as production systems: you will version instructions, tools, and evaluations; analyze traces and failure modes; measure quality, latency, and cost; and test every change against representative cases. You will adapt the speed of development and level of control to the degree of risk. In low-risk cases, you will experiment with a high level of autonomy. In more sensitive processes, you will introduce permissions, traceability, validations, action limits, data isolation, and human oversight. You will turn the lessons learned from each project into reusable components, evaluations, and patterns, helping make every new agent faster and more reliable to build than the previous one. The goal will not simply be to launch a demo, but to demonstrate that an agent can improve a real-world operation in a measurable, reliable, and responsible way. Eres lo que estamos buscando si tienes... A university degree in Engineering, Mathematics, Physics, or another related scientific, technical, or quantitative discipline. At least five years of experience in AI Engineering, Applied AI, Software Engineering, Product Engineering, Data, or a similar technical role. A strong programming foundation, particularly in Python. Experience building applications with language models and an understanding of concepts such as context engineering, tool calling, structured outputs, RAG, evaluations, and human-in-the-loop systems. Experience working with APIs, databases, authentication, testing, Git, and cloud-deployed services. The ability to turn an ambiguous problem into an architecture and an initial implementation without needing a fully finalized specification. The ability to use data, metrics, traces, and evaluations to analyze and improve the behavior of an AI system. Strong engineering judgment when balancing quality, latency, cost, security, maintainability, and development speed. An interest in working closely with users and the ability to communicate clearly in English with both technical and business stakeholders. Destacarás si tienes… Experience deploying agents or language-model-based products used by real users. Experience with tool calling, RAG, MCP servers, vector databases, workflows, or agent architectures. Experience building evaluation, observability, monitoring, or human-in-the-loop systems for AI applications. Knowledge of cloud infrastructure, containers, CI/CD, queues, or event-driven architectures. Experience using AI-assisted development tools to build, test, and maintain software. Experience working with sensitive data, access controls, traceability, or AI systems in regulated environments such as insurance or insurtech. ¿Por qué Tuio? At Tuio, we believe in end-to-end ownership. Everyone is responsible for understanding the problem, proposing a solution, building it, testing it with users, and measuring its results. We prioritize speed and learning, but we do not confuse speed with a lack of judgment. We accept technical debt when it helps us validate a hypothesis, and we know how to recognize when a solution needs to become more robust before its scope is expanded. Artificial intelligence is not simply another tool in our day-to-day work. It is a fundamental part of how we build products, operate, and improve the company. In addition, we offer: Competitive salary based on your experience and expertise. We want to attract the very best talent. Flexible working hours: we do not believe in enforcing a rigid 9:00 a.m. to 6:00 p.m. schedule. We believe in commitment, responsibility, and giving you the freedom to organize your work in the way that works best for you and the team. Remote work: although, should you feel like coming into the office, we are based in Madrid! A growth plan within a rapidly expanding company. The opportunity to work at a company where AI is not a trend, but a fundamental part of how we build products and improve our work every day. The opportunity to build solutions with real impact: you will work on real operations and use cases, not isolated demos. Top benefits: flexible compensation, health insurance, Tuio insurance policies, home-office equipment, team-building events—check out our LinkedIn profile!—and much more. Proceso de selección If you like what you have read, apply directly. Should you also like to tell us a little more about yourself, email us at . We will contact you so we can get to know each other. The process will be as follows: A conversation with our Head of AI & Data, who will explain the project, the team, and the challenges of the position. A video call with the People team—we promise not to ask you about your strengths and weaknesses. A conversation with our co-founder in charge of AI Deployment. A practical case in which you will need to understand an operation, identify an opportunity, and build or propose an initial agent-based solution. A review of the case with the AI & Data team, including the decisions you made, the experiments you carried out, the metrics you used, and the risks you identified. Welcome to the team!