Machine Learning Engineer – Compliance Model

We are looking for a Machine Learning Engineer to own the lifecycle of a jurisdiction-aware compliance model, from data preparation and supervised fine-tuning through evaluation, deployment, and production monitoring.

The model analyzes a document’s anchored metadata against a named standard and determines whether the document is compliant and what is missing. This role is focused on building a reliable, explainable model whose outputs can provide evidence to auditors and regulated customers.

Key Responsibilities

• Own the end-to-end ML lifecycle, including data, training, evaluation, deployment, and production behavior.
• Fine-tune and evaluate models for compliance classification and assessment.
• Design rigorous evaluation frameworks using held-out datasets, regression suites, and measurable acceptance criteria.
• Monitor model performance after deployment, identifying drift and regressions.
• Analyze model outputs and provide clear reasoning behind model behavior.
• Assess when a machine learning model is not the appropriate solution for a given problem.
• Work with the Google Vertex AI and Gemini model family environment.

Required Experience

• Hands-on experience with supervised fine-tuning of small- to mid-size models in production, beyond notebook-based experimentation.
• Strong experience designing rigorous ML evaluation strategies, including held-out sets, regression testing, and measurable acceptance criteria.
• Experience with model drift and regression monitoring after deployment.
• Strong understanding of model behavior, reliability, and explainability.
• Ability to work in environments where traceable and defensible model outputs are critical.

Work Conditions

• Work model: Remote LATAM
• Duration: Long-term
• Workload: Full-time
• Time zone: Requires 4 hours of overlap with US Eastern Time.
• Starting from: 16:00 Romania Time.