AI Platform Engineer

A-SAFE DIGITAL; Málaga; Híbrido; EUR 60.000 - 90.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

Career growth

Collaborative environment

Hybrid work

Pool & ping-pong

Summer hours (Aug) 8:00–15:00

Flexible Fridays

Descripción de la vacante

A-SAFE DIGITAL in Malaga is expanding its AI vertical and seeks an AI Platform Engineer to build production agents and own the MCP layer over core systems like NetSuite. You will enable secure, auditable AI adoption and guard against risks while collaborating with business stakeholders.

You will work in an AI-native stack with Claude Code, AWS, Kubernetes, Terraform, and CI/CD tooling, delivering measurable improvements and cost awareness across teams.

Formación

• Production-grade agentic systems experience, with multi-agent orchestration, retries and fault handling.
• Experience with MCP servers, tool registries and secure agent-to-data connections.
• Strong evaluation discipline with ability to explain a designed evaluation and its outcomes.
• Solid engineering foundation in Python and/or Ruby, APIs, cloud platforms and CI/CD; able to ship production code.

Responsabilidades

• Build production agents and multi-agent workflows that reason, call tools and recover from failures.
• Own MCP layer company-wide, architect and operate MCP servers over core systems.
• Create secure agent-to-data connections and safe, auditable data access.
• Drive AI adoption across teams through shared skills, standards and guardrails.
• Design evaluations, monitor latency, cost and reliability, and calibrate judgement.
• Set guardrails for prompt-injection defence, least privilege and governance for agents touching business data.
• Work in an AI-native environment using Claude Code, MCP OAuth, AWS, Kubernetes, Terraform, GitHub, Jira, Confluence, Sentry/tracing.

Conocimientos

Agentic systems

Multi-agent orchestration

Python

Ruby

CI/CD

Cloud platforms

Security-minded

Herramientas

MCP

OAuth/PKCE

AWS

Kubernetes

Terraform

GitHub

Jira

Confluence

Sentry

Descripción del empleo

AI Platform Engineer

Job title: AI Platform Engineer

Location: Malaga

Hours: 37.5 hours

Working pattern: Hybrid, 3 days per week in the office

Are you passionate about building production AI systems that make a real business faster and smarter? At A‑SAFE Digital in Malaga the company is creating a new AI vertical focused on practical, secure and scalable AI adoption across the group. As our AI Platform Engineer you will build production agents, own the MCP layer over our core systems, and help colleagues across the business use AI in a way that is trusted, measurable and safe.

How you will make an impact

• You will build production agents, designing multi‑agent workflows that reason, call tools, hold state and recover from failure.
• You will own the MCP layer company‑wide, architecting and operating MCP servers over core systems such as NetSuite, including tool registries, OAuth/PKCE, domain isolation, versioning and audit trails.
• You will connect company data safely, creating secure agent‑to‑data connections that allow agents to work with real business systems responsibly.
• You will drive AI adoption across teams, building shared skill libraries, standards such as CLAUDE.md / AGENTS.md, and practical guardrails that help non‑technical teams get value from AI.
• You will own evaluation and observability, designing evals, calibrating judgement, tracing non‑deterministic systems, and monitoring cost, latency and reliability.
• You will set the guardrails, thinking ahead about prompt‑injection defence, least‑privilege access, safety, governance and auditability for agents touching real business data.
• You will work in an AI‑native environment, using Claude Code, MCP OAuth, AWS, Kubernetes, Terraform, GitHub, Jira, Confluence, Sentry/tracing and our wider data and platform ecosystem.

What we are looking for

• Experience building agentic systems that have reached production, including multi‑agent orchestration, tool‑calling, state, retries and common failure modes.
• Hands‑on experience with MCP, including building or consuming servers, tool registries and secure agent‑to‑data connections.
• Strong evaluation discipline, with the ability to explain an evaluation you designed and what it showed.
• A strong engineering foundation in Python and/or Ruby, APIs, cloud platforms and CI/CD, with experience shipping production code.
• A security‑minded approach, including prompt‑injection defence, least privilege, governance and auditability.
• Confidence working with business stakeholders and bringing non‑technical teams along, because adoption is a key part of this role.
• An interest in measurable outcomes, using DORA and AI productivity metrics rather than assumptions.

Nice‑to‑haves

• Depth in MCP and/or A2A protocols, agent UX, latency optimisation and cost optimisation.
• Experience rolling out AI across a whole organisation, not just within one engineering team.
• Familiarity with the agentic AI ecosystem, such as Claude Agent SDK, LangGraph, PydanticAI and observability tooling such as Langfuse.
• ERP or CRM integration experience, particularly with NetSuite.
• Exposure to AWS Bedrock, EKS, NVIDIA DGX, PostgreSQL, ClickHouse, CDC pipelines, Kubernetes, Terraform or similar technologies.

Benefits

• Opportunities for professional growth and career advancement.
• Collaborative and supportive work environment.
• Hybrid model: 3 days in the office, 2 days remote work.
• On‑site pool and ping‑pong table.
• Intensive working hours (Jornada intensiva) in August: during this month we work from 8:00 to 15:00.
• Flexible working hours with the option to finish early on Fridays.