описание
Generative AI is no longer just about prototypes. The challenge is to build products and systems that work reliably in production, are secure and traceable, and deliver value where decisions matter. At NTT DATA, we are looking for someone with architectural vision, technical judgment, and a passion for leadership to drive the end-to-end development of agentic AI solutions. A role for someone who wants to stay close to the code while guiding a team in adopting practices that are redefining how software is created: Spec-Driven Development, AI agents, MCP, skills, and context engineering.
задачи
- Technically lead an AI engineering and full-stack development team, serving as a reference for architecture, quality, and good practices.
- Design solutions based on LLMs, workflows, agents, and multi-agent systems, defining the appropriate level of autonomy for each use case.
- Implement Spec-Driven Development with OpenSpec, GitHub Spec Kit, or equivalent tools, ensuring clear specifications, traceability, and validation.
- Build full-stack applications with agents, MCP servers, skills, and context areas, participating directly in development.
- Establish standards for testing, LLM evaluation, observability, cost control, and quality for code developed by people and agents.
- Design security and governance mechanisms: guardrails, least-privilege permissions, human-in-the-loop, secret management, and alignment with GDPR and the AI Act.
- Translate business needs into understandable technical decisions, connecting architecture, feasibility, and value.
- Support the team's technical growth through mentoring, design reviews, and shared learning.
требования
- More than 10 years of experience in software engineering, with at least 3 years technically leading teams.
- Recent experience — at least 2–3 years — building and deploying generative or agentic AI solutions in production.
- Expert command of Python and TypeScript.
- Practical experience with Spec-Driven Development, using OpenSpec, GitHub Spec Kit, or equivalent approaches based on verifiable requirements.
- Strong knowledge of agentic AI architectures and frameworks, such as LangGraph, Claude Agent SDK, OpenAI Agents SDK, Google ADK, or CrewAI.
- Ability to design and develop MCP servers and clients, including tools, resources, and authentication.
- Experience building APIs, frontend applications, integrations, test automation, and cloud deployments.
- C1 level of Spanish and English, with the ability to communicate technical decisions clearly to both technical and business audiences.
- Будет плюсом: Experience creating internal AI-assisted development platforms or agent-based software factories; Open-source contributions, publications, or participation in communities focused on AI, SDD, or MCP; Experience in insurance, banking, healthcare, public administration, or other regulated environments; Knowledge of open models, local deployment, vLLM, Ollama, or fine-tuning; Experience with AWS, Azure, or Google Cloud, CI/CD, containers, AI Gateway, Langfuse, LangSmith, or equivalent solutions
условия
- A technical challenge with impact: taking agentic AI from architecture to real-world operations, with standards for quality, security, and traceability.
- Room to influence how software is built: defining practices, patterns, and tools that scale beyond a single project.
- Ongoing professional development, learning alongside specialists, and access to high-level technology initiatives.
- A hybrid working model from Barcelona, with flexibility and a trust-based way of collaborating.
- Social benefits, wellbeing programs, and work-life balance measures designed to support you inside and outside work.
- An environment where technical ambition goes hand in hand with closeness, respect, and collaboration between people.