Machine Learning Engineer

About CarOnSale

CarOnSale is a leading AI-driven platform dedicated to B2B used car trading across Europe. With more than 40,000 buyers spanning over 20 countries, and 85% exclusive inventory, we integrate software, pricing intelligence, logistics, and financing into a unified operating system for the industry.

Role Overview

Currently, our production environment supports four machine learning models, with plans to scale up to 15–20 by mid-2027. This growth depends on a robust shared pipeline encompassing data extraction, validation, transformation, training, and evaluation stages. You will be responsible for owning everything after model handoff, including packaging, deployment, monitoring for model drift, and approving models for production service.

Responsibilities

• Manage models from handoff through deployment, ensuring they are properly packaged, deployed, and monitored for reliable performance.
• Detect model drift and maintain performance monitoring with effective alerting and incident response procedures.
• Oversee serving and inference pipelines, including handling fitted pipeline artifacts, inference endpoints, monitoring integration, and ensuring parity with feature stores.
• Conduct thorough reviews of model design and evaluation methodologies to identify issues such as data leakage or evaluation errors early in development.
• Develop and enhance the shared ML platform to maintain its integrity without introducing project-specific logic.
• Establish and enforce engineering standards as the platform scales throughout the organization.

Candidate Qualifications

• Minimum of two years of experience in production machine learning engineering with accountability for models post-handoff.
• Proficient in writing typed, tested, production-quality Python code and reviewing others' code.
• Deep understanding of machine learning concepts to critically evaluate pipeline problem framing, feature engineering, model selection, and evaluation.
• Hands-on expertise with managed ML platforms such as SageMaker, Vertex AI, Databricks, or Azure ML; familiar with feature stores, CI/CD for ML, AWS, and Terraform.
• Comfortable using AI-assisted development tools like Claude, ChatGPT, or Copilot daily.
• Fluent English at a C1 level, both written and spoken; German language proficiency is not required as English is the working language.

Preferred Skills

• Experience with Snowflake and dbt is a plus, with willingness to learn if unfamiliar.
• Previous experience mentoring team members or conducting code reviews.
• Aptitude for operating in ambiguous situations where solutions are not predefined.

Work Arrangement

This position is based in Berlin Schöneberg, working in a hybrid setup with three days onsite at the office and two days remote work.

Benefits

• Hybrid work model: 3 days in the office, 2 days remote, plus 25 "Work from Anywhere" days annually.
• 28 days paid annual leave.
• Biannual career and development discussions.
• Company pension plan with 20% employer contribution.
• Fully covered public transit ticket (Deutschlandticket).
• Subsidies for FitX gym membership or Urban Sports Club.
• Virtual stock options allowing participation in company growth.
• Modern IT tools and infrastructure for daily work.
• Comprehensive onboarding program with buddy system and social events.
• Diversity and inclusion initiatives including active women's network, meditation and prayer room, and a pet-friendly office environment.