Machine Learning Engineer

Responsibilities

• Design, build and deploy production machine learning systems that support newsroom, editorial and new-audience initiatives.
• Develop classification, prediction and ranking models using proprietary newsroom data and editorial expertise — a modeling foundation that extends across new sector products (Decoded, Code & Counsel, Energy, Pharma) rather than requiring a bespoke build per vertical.
• Build datasets, feature engineering pipelines and evaluation frameworks to improve AI system performance, including turning the Leadership team's data-needs definitions into working systems.
• Experiment with prompting, retrieval, fine-tuning and other model adaptation techniques to solve journalism-specific problems.
• Partner closely with journalists and editors to translate editorial knowledge into machine learning features, labels and evaluation criteria.
• Collaborate with software engineers to integrate machine learning capabilities into production applications.
• Measure and improve model quality using rigorous evaluation methodologies that balance accuracy, editorial standards and operational performance.
• Evaluate new AI technologies while considering performance, cost, maintainability and long-term flexibility.
• Building systems that classify the newsworthiness of stories using editorial judgment.
• Developing prediction models that combine public information with POLITICO's proprietary reporting expertise — the same approach that powers POLITICO Signals and Agency IQ.
• Improving AI-assisted editing by helping models better understand and apply POLITICO's editorial style guide.
• Creating proprietary classifiers that enrich public datasets with editorial intelligence, extendable to new sector products as they launch.
• Designing evaluation frameworks to measure AI quality before newsroom deployment.
• Optimizing AI workflows to improve performance while reducing inference costs.
• Supporting investigative and enterprise reporting through machine-learning-assisted analysis of complex datasets.
• Improve the accuracy and reliability of AI systems used by POLITICO's newsroom.
• Build evaluation frameworks that help teams confidently assess AI quality and editorial alignment.
• Contribute to proprietary classifiers, prediction systems and editorial intelligence capabilities that differentiate a named initiative — a sector product, POLITICO Signals or Agency IQ.
• Partner effectively with software engineers, product teams and newsroom stakeholders to bring machine learning solutions into production.
• Help establish a strong technical foundation for future AI initiatives across the newsroom and our new lines of business.

Requirements

• 5+ years of professional experience building or deploying machine learning systems.
• Strong software engineering skills, preferably in Python.
• Experience with machine learning fundamentals including supervised learning, feature engineering, model evaluation and experimentation.
• Experience with modern machine learning frameworks such as PyTorch or TensorFlow.
• Familiarity with large language models, embeddings or natural language processing techniques.
• Experience working with structured and unstructured data.
• Strong problem-solving and analytical skills.
• Excellent communication skills and the ability to collaborate with engineers, product managers and non-technical stakeholders.

Qualifications

• Experience adapting or fine-tuning foundation models.
• Experience building NLP, information retrieval, classification or ranking systems.
• Familiarity with model evaluation frameworks, experiment tracking or MLOps practices.
• Experience deploying machine learning models to cloud infrastructure.
• Interest in journalism, media, public policy or knowledge-intensive domains.

Pay

• $175,000 - $225,000 Annual Base. The final salary offered will be based on factors including relevant experience, skills, and qualifications.