Senior Data Scientist — LLM/ML Production Lead

Moss; Amsterdam; On-site; EUR 120,000 - 180,000; Full time

Benefits offered by this job

Equity

In-person culture

Learning budget

Job summary

Moss is expanding its Data & AI team in Amsterdam and seeks a senior/lead contributor with hands-on LLM app experience. You’ll own end-to-end LLM/ML initiatives—framing problems, modeling, evaluating, deploying, and iterating to improve core workflows like invoice understanding, approval matching, reconciliation, anomaly detection, and user assistance.

The role requires 5–10+ years in applied ML with production ownership, strong Python and ML library skills, and the ability to balance model

Qualifications

• 5–10+ years in applied ML with production ownership of shipping models.
• Hands-on LLM app/agent experience (tool use, function calling, RAG, structured outputs) in production.
• Strong Python + ML stack (scikit-learn, XGBoost/LightGBM, PyTorch/TF) and robust experiment/evaluation skills.
• Strategic Problem Solver: approach ambiguous problems by deeply understanding business context and framing data science solutions that drive meaningful outcomes.
• End‑to‑End Executor: independently own projects from ideation to production, balancing model performance with engineering realities.

Responsibilities

• Own LLM/ML pipelines for extraction, classification, anomaly detection, and recommendations.
• Define evaluation & monitoring: gold sets, automated tests, robustness checks, SLAs, and cost/latency tracking.
• Productionize with Platform/Data Eng: bring your models live and iterate/improve until they are adopted by your stakeholders.

Skills

Applied ML with production ownership

Hands-on LLM app/agent experience

Python + ML stack

Strategic problem solving

End-to-End project ownership

Tools

scikit-learn

XGBoost

LightGBM

PyTorch

TensorFlow

Job description

Moss is expanding its Data & AI team in Amsterdam and seeks a senior/lead contributor with hands-on LLM app experience. You’ll own end-to-end LLM/ML initiatives—framing problems, modeling, evaluating, deploying, and iterating to improve core workflows like invoice understanding, approval matching, reconciliation, anomaly detection, and user assistance.

The role requires 5–10+ years in applied ML with production ownership, strong Python and ML library skills, and the ability to balance model