Role Overview
RiskMaat is building a Reasoning Machine for systemic risk. The Machine Learning Engineer will develop learned models that transform diverse documents into structured knowledge, enabling reliable decision-making. You will work across language models, graph learning, and deterministic methods to address complex challenges.
Responsibilities
- Convert diverse documents into structured knowledge about organizations and dependencies.
- Build and train graph neural networks and other learned models.
- Develop methods to keep the model current, including confidence estimates.
- Create evaluation frameworks and automated checks for model quality.
- Collaborate across language models, graph learning, and deterministic methods.
Requirements
- Master's degree or equivalent in computer science, machine learning, or a related field.
- At least one year of experience building data or AI systems for production.
- Excellent Python skills: clean, typed, tested code.
- Experience with language model pipelines, including structured outputs and evaluation.
- Experience with graph neural networks and representation learning.
- Fluent English.
Nice to Have
- Temporal or dynamic graph models and link prediction.
- Uncertainty estimation and calibration.
- Information extraction in finance or cybersecurity.
- Experience combining deterministic evaluators with language model components.
- Knowledge of data contracts and schema design.
Benefits
- Frontier engineering opportunity.
- Ownership from day one.
- Research-driven environment.
- Transparent compensation (TVöD E 13, Stufe 2).
- Munich-based team.