AI Data Scientist

You will design, build, and run AI systems for our clients across Belgium.

Hands-on building, in production, at scale.

What will you do?

• Build end-to-end RAG solutions over enterprise data.
• Integrate Model Context Protocol (MCP) to connect LLMs with internal tools and data.
• Design multi-agent architectures to automate complex workflows.
• Fine-tune and evaluate LLMs; implement guardrails for hallucination, toxicity, and relevance.
• Deploy to production: API, monitoring, scaling.

The Technical Stack

• LLMs: OpenAI GPT-4/4o, Azure OpenAI, Claude, LLaMA, Mistral, HuggingFace.
• RAG: LangChain, LlamaIndex, vector stores (Pinecone, Weaviate, Chroma), Azure AI Search.
• MCP / agents: Anthropic MCP, LangGraph, Semantic Kernel, AutoGen, CrewAI.
• Classic ML: Scikit-learn, XGBoost, PyTorch, TensorFlow, MLflow.
• Infra / deploy: Azure ML, Docker, Kubernetes, FastAPI, Streamlit.
• Languages: Python (primary), SQL, TypeScript.

The Mindset we are looking for:

• You ship production systems, not POCs.
• You adapt your communication to your audience.
• You stay rigorous: you don't overclaim.
• You take ownership end to end.
• You design for trust: evaluation and guardrails from the start.
• You mentor through how you work.

3-8 years in Data Science, including 1-3 in Artificial Intelligence/LLM.

Mid to senior level.

Right to work in Belgium and Brussels mobility required.

Languages

Fluency in English is required, along with full professional proficiency in either French or Dutch.

• Production Generative AI deployments you can point to.
• A Master's/PhD in CS, mathematics, physics, or engineering.
• Azure AI Engineer or Azure Data Scientist Associate certification.
• Public work: open source, papers, conference talks.
• Experience in a regulated vertical: finance, pharma, public sector.

What Exsolvæ Offer:

• Transparent compensation, discussed from the first interview, plus our Solver Incentive Program.
• Full focus on the work, no internal reporting overhead.
• Engagements we can deliver well.
• Real growth paths and time to publish or speak.

What this means for you:

• Work on production Artificial Intelligence (RAG, agents and MCP). For clients serious about deploying it.
• A €2,000 annual Knowledge & Networking budget for certifications and conferences.
• A GenAI workspace and current tools from day one, through our OpenAI partnership.
• Titles earned through demonstrated work.