8 months
23/04/2026 - 31/12/2026 Brussels, Belgium
Requirements
Roles
  • Artificial Intelligence (AI) Engineer Senior
Languages
  • Dutch Native or bilingual proficiency
  • French Native or bilingual proficiency
  • English Full professional proficiency
Skills
  • Python Advanced
  • Vector databases Intermediate
  • Terraform Advanced
  • LLM / Large Language Models Intermediate
  • Azure Kubernetes Service - AKS Advanced
  • Azure OpenAI Service Intermediate
  • RAG (Retrieval-Augmented Generation) Intermediate
  • Azure DevOps Advanced
Description

ProUnity est le point de contact unique (SPOC) pour cette mission. Si vous avez des questions, vous pouvez contacter Musli Muslija, Client Satisfaction Manager, par e-mail à paradigm@pro-unity.com

Job Description

Context

STIB-MIVB (Brussels public transport operator) is building UGo AI Search, a RAG-based bilingual French/Dutch knowledge search service serving ~10,000 employees across the organization. UGo AI Search is the first production service on the STIB AI Platform and will be followed by additional AI services.

Your role

We are recruiting a Senior AI Engineer to join the Digital Innovation / AI CoE team for the Build and early Run phases of UGo AI Search. You will work alongside STIB’s internal AI Platform Architect and UGo AI Search Solution Architect, turning the High-Level Design into a running, evaluated, and maintainable service.

The AI Platform runs on a hybrid operating model: managed Azure services (Azure OpenAI, Azure API Management, Azure AI Content Safety) combined with self-hosted open-source components on AKS (vector store, orchestration, ingestion pipeline). The Digital Innovation / AI CoE team owns Infrastructure-as-Code for all application-scoped Azure resources, using Terraform and Azure DevOps pipelines.

What you will deliver

You will implement the RAG pipeline end-to-end (retrieval, reranking, LLM orchestration, citation handling), the ingestion pipeline (Jahia CMS and SharePoint connectors, chunking, embedding), security trimming and ACL propagation, prompt templates and guardrails, the evaluation harness, and the Terraform and Helm artefacts needed to deploy and operate the service. You will also contribute to running and tuning the service during the early Run phase.

Informations additionnelles:La mission peut être reconduite pour une durée maximale (durée initiale comprise) de : 880 jours ouvrables.