Senior AI Engineer

This is what we offer you

• Gross monthly salary between €5,381 and €7,687 (scale 10)
• Thirteenth month's salary and 8% holiday allowance
• 10% Employee Benefit Budget
• EUR 1,400 development budget per year
• Hybrid working: balance between home and office work (possible for most roles)
• A pension, for which you can set the maximum amount of your personal contribution

Onboarding, extracting, transforming, and loading (un)structured and normalized information into our monitoring solutions (Splunk, Microsoft Sentinel and more) based on input from the system and application owners is the main part of your job. Together with the engineers for security-, application & logfile monitoring you will deliver high value monitoring that is used by the Rabobank Control Room, SOC and application teams.

The team is facing an interesting but challenging project to migrate multiple data integration solutions into one. This project concerns migrating from ArcSight connectors, Logstash and Splunk towards Cribl. We are now at the starting point of implementing our new Cribl environment and designing our migrations for the current solutions. It is your role to highly participate in this with your experience in data integration and create synergy within the team to get every engineer on the same level.

We are seeking a skilled Prompt Engineer with a strong background in large language models (LLMs) and a passion for advancing AI performance through systematic prompt design, fine-tuning, and evaluation. You should have a hands-on understanding of both model adaptation techniques and cost-performance optimization.

Top 3 responsibilities

• Develop and fine-tune prompt strategies to accurately interpret and parse various log types (e.g., syslog, JSON, app logs). Create effective few-shot examples and systematic variant testing to ensure high-fidelity data extraction and downstream parser generation.
• Develop health checks for log sources, parsing accuracy, ETL workflows, and parser output quality. Use LLM-driven metrics (e.g., token confidence, completeness, schema match rate) to monitor and continuously improve data integrity.
• Use QLoRA or reinforcement learning techniques to adapt models for parsing edge cases. Monitor performance metrics (latency, token usage, confidence) and refine prompts/models to ensure scalable, cost-effective parsing at production scale.

You and your talent

• 2+ years of experience in prompt engineering for LLMs.
• Solid foundation in LLM fine-tuning, particularly using QLoRA and reinforcement learning-based adaptation methods.
• Proven ability to design effective few-shot learning examples that maximize model performance.
• Hands-on experience with prompt variant generation and A/B testing strategies.
• Strong grasp of probabilistic confidence metrics and how to leverage them in prompt evaluation.
• Experience building or contributing to automated prompt optimization systems.
• Expertise in Retrieval-Augmented Generation (RAG) or Contextual RAG (CRAG) pipelines.
• Demonstrated ability to systematically monitor, evaluate, and fine-tune LLMs with cost-efficiency in mind.
• Proficiency in JavaScript and experience fine-tuning models using Azure OpenAI Service.

Together we achieve more than alone

We believe in the power of difference. Bringing together people's differences is what makes us an even better bank. So, we are very curious about what you can bring to our Security Data Integration team at Tech4Ops.

The data integration team is a group of open-minded people, that are always ready to support each other and living the Rabobank value: “I make you better!”. Our team has a strong focus on innovation, R&D and data quality in everything we do
Albert Wollaars, IT Lead Tech4Ops.

To be considered for this position, you must be located in the Netherlands or elsewhere in the EU and have the legal right to work in the Netherlands. We are unable to consider candidates requiring relocation from another country.