DevOps Engineer

For Rabobank in Utrecht, we are looking for an AI Engineer for GenAI Assistants within Customer Service.

Important:

• We are looking for 1 FTE;
• We will close the vacancy before the indicated deadline when we have received enough professionals;
• Please clearly indicate in the motivation field whether the candidate has any planned vacation or special leave during the duration of this assignment;
• Please provide in the CV whether an internal reference can be requested. If the candidate is currently working at, or has recently worked at Rabobank, we would appreciate receiving the name and job title of the relevant team lead;
• Language: English and Dutch mandatory;
• ZZP Allowed: No.

About us & the role:

We believe customer interactions are evolving toward AI agents that actively support personalized financial choices. These agents depend on high-quality enterprise data and knowledge to provide accurate, grounded, and reliable support. Preparing this information and systematically evaluating the quality of AI assistants are essential to creating valuable customer-service solutions.

That is why we are looking for an AI Engineer to strengthen our Conversational AI teams. You will focus on preparing enterprise data and knowledge for advanced AI assistants and applying evaluation frameworks across single-turn, multi-turn, simulated, and agentic scenarios. In this role, you will help ensure that our assistants provide accurate, grounded, and reliable support to advisors.

You will work in a hybrid setup, ideally spending one day a week onsite in Utrecht, collaborating closely with teams to implement new conversational and AI concepts. On Mondays, our department works from the office. For the rest of the week, we coordinate among ourselves whether we work from home or the office. The Digital & Customer Interaction Tribe aims to deliver an excellent customer experience, regardless of how the customer contacts Rabobank. Core values of our area include fun, collaboration, proactivity, and problem-solving.

Your main focus will be on:

• Collaborate within a scrum team to prepare, structure, enrich, and validate enterprise data for AI knowledge bases and AI Chatbots;
• Build and maintain data pipelines in Databricks to process structured and unstructured data for AI knowledge bases;
• Apply evaluation frameworks to test advanced chatbots across single-turn, multi-turn, simulation-based, and agentic scenarios;
• Prepare, clean, transform, and enrich source data to improve the quality and reliability of information used by AI assistants;
• Create and maintain evaluation datasets covering realistic conversations, complex cases, edge cases, and multi-step workflows;
• Implement automated testing, monitoring, and quality gates for changes to data, prompts, models, retrieval, and orchestration;
• Deploy and manage cloud infrastructure using Infrastructure as Code, particularly Azure Bicep and CI/CD pipelines;
• Analyse evaluation results and translate findings into improvements across knowledge, retrieval, prompts, tools, and application logic.

Your Talent:

• A strong engineering and data-quality mindset, with an affinity for Agile development;
• A proactive, hands-on approach to solving complex AI-quality and cloud-engineering challenges;
• Strong critical thinking, ownership, and accountability for production-ready solutions;
• Excellent communication skills and a customer-centric attitude;
• A relevant HBO or university degree and fluency in English.

Your Skillset:

• Strong proficiency in Python and experience with data engineering and automated testing;
• Hands-on experience with Azure, Databricks, DevOps, CI/CD, and Infrastructure as Code, preferably Bicep;
• Experience building data pipelines and preparing structured and unstructured enterprise data for AI knowledge bases and RAG applications;
• Experience applying evaluation methods to advanced chatbots, including multi-turn, simulated, and agentic interactions;
• Familiarity with Context Engineering, Agentic AI, Function Calling, retrieval strategies, and tool use.

We would like to receive in the personal motivation an explanation showing that the candidate possesses the required competencies.

Additional information:

• Work is done in a hybrid way, according to the Rabobank @anywhere principle;
• Due to increasing congestion at the Rabobank location on Croeselaan in Utrecht, a new parking policy will take effect from April 2026. From Monday to Thursday, external employees will no longer be able to park on the Croeselaan site. Only on Fridays will parking for externals still be possible;
• Please note: employment-freelance constructions are not allowed according to Rabobank policy. Do not submit candidates through this construction. Compliance will also be checked during screening;
• Candidate suppliers must be aware of the applicable laws and regulations regarding employment conditions and the Rabobank collective labor agreement. This role falls under scale 09 of the Rabobank CAO;
• We would like to receive the personal motivation and CV in Dutch;
• The candidate may be submitted for only one application at a time. This means that during the exclusivity period of 4 working days, the candidate can only be submitted to Magnit - Rabobank;
• It is also important that the candidate is available for the entire period requested in the application;
• Pre-employment screening: If the candidate is allowed to start with the client, a pre-employment screening will take place. As of July 1, 2026, a new PES arrangement will apply. The PES is carried out by an external party. More information can be found in Annex A. Candidates can only start after completing the pre-employment screening;
• The candidate has a valid identification document, a passport or ID card, which must be valid both at intake and on the contract start date, and can be presented for verification;
• When proposing a freelancer indirectly, the contract will be handled via a fee-only agreement. This agreement can be requested in advance;
• When submitting a candidate, we assume that you agree with the terms of this specific client. If you are not familiar with these terms, you can request them from the responsible recruiter.

IMPORTANT!:

Fill in all candidate information correctly. Enter the candidate’s contact information under “email address and phone number”! The same goes for the address details. These are the details the client wants to receive with the submission. Incomplete or incorrect submissions may result in candidates not being considered and will lead to a rejection.