A major public transport organisation is building a bilingual French/Dutch retrieval-augmented generation (RAG) knowledge search service that will be the first production workload on its AI platform and will serve about 10,000 employees. The Senior AI Engineer will implement the service on Azure, combining managed services such as Azure OpenAI Service with self-hosted components on Azure Kubernetes Service (AKS), and will use Terraform, Azure DevOps and Python to deliver a production-ready RAG application.
The mission
The project delivers a RAG-based bilingual knowledge search service that indexes corporate content (Jahia CMS, SharePoint and other sources), embeds passages into a vector store, and orchestrates LLM calls with reranking, citation handling and content-safety checks. The platform is a hybrid model using Azure managed services (Azure OpenAI, API Management, Content Safety) alongside self-hosted vector and orchestration components on AKS. The service will support roughly 10,000 internal users and is intended to set the standards for subsequent AI services on the platform.
You will join the Digital Innovation / AI CoE working directly with the platform architect and the solution architect to turn the high-level design into a running, evaluated, and maintainable service. Day to day you will implement and deploy the RAG pipeline, build ingestion connectors for content sources, author Terraform and Helm artifacts for deployment, and contribute to the early Run phase by monitoring performance, tuning prompts and models, and improving the evaluation harness.
Your responsibilities
- Implement the RAG pipeline end to end, delivering retrieval, reranking, LLM orchestration, citation handling and evaluation metrics so the service can run in production.
- Build and maintain ingestion connectors, including Jahia CMS and SharePoint connectors, chunking, embedding and vector storage so content is searchable and auditable.
- Deliver Infrastructure-as-Code and deployment artifacts, authoring Terraform modules and Helm charts and integrating them into Azure DevOps pipelines for repeatable deployments.
- Integrate and operate Azure managed services such as Azure OpenAI Service, API Management and Content Safety with AKS-hosted components to meet reliability and security objectives.
- Implement security trimming and ACL propagation so search results respect user permissions and organisational policies.
- Run the early production phase: collect evaluation data, tune prompt templates and guardrails, and improve model selection and routing based on measured metrics.
Your profile
Essential skills
- Proven experience as a Senior AI Engineer or similar role delivering LLM-based services, with hands-on work on RAG systems and LLM orchestration.
- Strong Azure experience, specifically with Azure OpenAI Service, Azure Kubernetes Service (AKS) and Azure DevOps for CI/CD.
- Practical expertise in Python for ingestion, embeddings and orchestration code.
- Experience with Infrastructure-as-Code using Terraform and packaging deployments with Helm.
- Familiarity with vector databases / vector stores and embeddings workflows.
- Ability to implement security trimming, ACL propagation and to design evaluation metrics for conversational / retrieval systems.