Job title: AI Platform Developer

Location: Brussels - 1 Day Per week

Engagement: Freelance

Duration: Until 31 December 2027 (extension possible)

About the role

Our client is expanding their online self-service platform with a Virtual Assistant. Working in an Agile team alongside architects you'll design and build the MCP layer that safely connects LLMs to the clients business capabilities and backend systems.

Day to day requirements

- Design, build and maintain MCP servers using Quarkus, keeping them scalable, performant and maintainable.
- Surface capabilities as MCP Tools, Resources and Prompts that agents can safely consume.
- Connect to existing APIs, DataHub APIs and backend systems.
- Contribute to reusable MCP frameworks, SDKs and standards the wider platform can build on.
- Implement secure authentication and authorization flows across the platform.
- Ensure auditability, observability and governance of every AI interaction.
- Support the evolution from single-agent to multi-agent AI scenarios.




- Help define the AI platform architecture and the best practices that go with it.
- Perform performance optimization, monitoring and troubleshooting.
- Turn requirements into working software and take responsibility for implementing the requirements.

Required Skills

- Java 17+ and Quarkus
- REST APIs, Dependency Injection and reactive programming
- MCP architecture, LLM Architecture, Tools, Resources, Prompts, MCP client/server interaction
- AWS Bedrock and Bedrock AgentCore
- LLM fundamentals, prompt engineering
- Tool calling, AI orchestration, context management, RAG
- Bedrock Agents / AgentCore, embeddings, AI model integration
- OAuth2 / OIDC, Entra ID integration, RBAC
- Secure API design, audit logging, threat modelling, secure software development

Please apply for more information.

📌 Ai platform developer (España)
🏢 La Fosse
📍 España