AI Agent Engineer

AI Agent Engineer — You will join a growing AI Team in a fast-moving fashion retailer and report to the AI Team Lead.

This is a hybrid role, partly from home and partly at our office in Hilversum, the Netherlands, so we are only considering applicants who live within commuting distance of Hilversum. We value demonstrated skill over a traditional career path, and the selection process includes a short practical exercise.

• 40 uur
• IT

This position is for an AI Agent Engineer to build and run AI agents that take on the routine work in Hunkemöller's head office roles. When a role opens, you work with the department to map its work, judge honestly how much an agent can take on, and build and run the agent that does it, up to one a month. You will work across Microsoft 365, Google Cloud and Anthropic, design agents that are safe to rely on, and help teams adopt them.

Responsibilities:

• Assess Automation Opportunities: Map a role's processes with the department and decide quickly whether an agent will hold up in daily use, including when the answer is no.

• Build Production Agents: Build agents in Microsoft 365, on Google Cloud or with Anthropic models that work across the same tools and data the role uses today.

• Design for Safe Use: Give every agent a named owner, read access by default, write access only with approval, and a person who confirms consequential actions.

• Run What You Build: Own monitoring, evaluation, maintenance and fixes for the agents you deliver.

• Support the Teams Using Agents: Onboard and train the people who work alongside each agent, and act on their feedback.

• Reuse and Document: Turn common patterns into reusable building blocks and document each agent so others can run it.

• Cover for the Team: Back up the Creative AI Engineer on basic work, so the team keeps moving when someone is out.

Requirements:

• You have hands-on experience building automation, such as robotic process automation, Power Platform or workflow tools, and recent work on agents built on language models. Recent, relevant experience matters more than years in a role.

• You have experience using Python or TypeScript, APIs and system integration, and experience with Microsoft 365 and Copilot Studio, Google Cloud agent tooling, or Anthropic models and agent tooling. Experience with the Model Context Protocol is a plus.

• Experience with testing an agent's output, handling exceptions and keeping a person in the loop where it matters.

• You know how to break a role into its tasks, spot what is routine and what needs judgment, and estimate the effort honestly.

• Have operated systems that people depend on, not only demos.

• You communicate clearly and work patiently with teams adjusting to new ways of working, as well as knowing how to build trust with HR and department leads.

• Clear understanding of data access and privacy, and the risks of agents that can change business data.

• Written and verbal English communication skills. Dutch is a plus.

What we offer

• 25 days of annual leave, with the option to buy or sell up to 4 additional days. Every year in April, you will be eligible to buy or sell additional vacation days.

• Hybrid work model, combining office and remote working. We have a 60:40 policy based on your working hours.

• Possibility to work from abroad for up to 2 weeks per year.

• An international work environment, working with teams across different countries.

• Travel allowance to support commuting costs.

• Access to the Hunkemöller Academy for professional development. Continuous learning opportunities and professional growth are supported through various training programs and workshops.

• 25% employee discount on all Hunkemöller products. On some Mondays we have an additional sale on our newest collections. On top of the 25% discount, you will receive an additional 25 - 35% discount.

• You will get 25 holidays per annum, based on a fulltime role.