Job Description

📋 Tasks and responsibilities

For a public transport organisation in Brussels, an AI Centre of Excellence is building several AI services for employees and customers. These include retrieval-based assistants and AI agents that call tools and business systems on behalf of users. The services run on a dedicated AI platform. The first services are scheduled to go live from December 2026, with further services planned for 2027.

Your role

As a Medior AI Engineer, you take ownership of the Run phase for AI services in production. You take over each service from the senior engineers who built it and keep it reliable, measurable, secure and cost-efficient over time.

You work with an AI platform architect, two senior AI engineers responsible for the Build phase, the business and content owners of each service, and the IT operations teams.

The platform follows a hybrid operating model. It combines managed Azure services through Microsoft Foundry and Azure OpenAI Service for LLMs and embeddings, Azure AI Search, Azure API Management and Azure AI Content Safety with self-hosted open-source components on Azure Kubernetes Service (AKS). These components include Python-based agent orchestration with LangChain and an ingestion pipeline. MLflow on Databricks is used for evaluation and tracing with OpenTelemetry. Infrastructure-as-Code for application-scoped Azure resources is managed with Terraform and Azure DevOps pipelines.

What you will deliver

Your activities will vary from service to service. Typical responsibilities include:

• Operate AI services in production by monitoring availability, latency and errors, handling incidents and service requests, analysing root causes, and fixing or escalating issues.
• Maintain answer quality by running evaluation sets, or golden sets, with business content owners. Track answer quality, non-answer rates and user feedback, and investigate quality regressions.
• Operate ingestion pipelines, including incremental refreshes, failed documents, index health, and the onboarding of new content sources or business domains within existing connectors.
• Monitor agent integrations with business systems, handle failures and timeouts, and review action logs and exceptions for agents acting on behalf of users.
• Manage changes safely, including model version upgrades and retirements, prompt and configuration changes, dependency updates and security updates. Validate each change against the evaluation sets before release.
• Track token consumption and the cost per service, set up alerts for cost drift, and propose optimisations.
• Maintain Terraform, Helm and Azure DevOps artefacts, and keep the DEV, TST, ACC and PROD environments consistent.
• Apply data protection rules for trace retention, erasure and access restrictions as defined in each service's data protection impact assessment.
• Maintain runbooks and documentation, and report service health regularly to service owners and the AI Centre of Excellence.
• Deliver small enhancements and fixes. Structural changes and new features remain with the Build team.

📝 Your profile

You have experience as a Medior Artificial Intelligence (AI) Engineer and are familiar with operating AI services in production.

Skills

• Azure DevOps
• Azure Kubernetes Service (AKS)
• Azure OpenAI Service
• LangChain
• LLM / Large Language Models
• Python
• RAG (Retrieval-Augmented Generation)
• Terraform
• Vector databases

Languages

• Dutch or French
• English

💼 Offer

You will be part of a growing Belgian SME where initiative and personal development are encouraged. We will provide you with an enjoyable work environment with fun colleagues. We will work out a career plan with you, with attention and a budget for extra education/certification. You can count on an attractive salary, supplemented with extra-legal benefits, including a company car.

(Freelance is also possible)