Software Engineer – Machine Learning (AI Training)

Machine learning code evaluation

Review AI-generated machine learning solutions for correctness and quality, contributing practical engineering judgment from Warsaw or elsewhere in Poland. This is a fully remote hourly contract with flexible, project-based commitments.

Work is asynchronous on your own schedule, with no fixed hours or meetings. There is no minimum hour commitment beyond what individual projects require.

Responsibilities

• Evaluate generated Python, TensorFlow, PyTorch, and scikit-learn code for correctness, efficiency, scalability, and clarity.
• Write ML solutions covering modeling, data processing, and deployment problems at different difficulty levels.
• Explain architecture choices, code logic, and problem-solving approaches in clear language for developers.
• Identify edge cases, ambiguities, and weaknesses in problem statements, datasets, and generated responses.
• Help define quality expectations for how AI understands and produces machine learning code.

Qualifications

• Strong practical knowledge of model development, data preprocessing, training pipelines, and deployment.
• Hands-on experience with TensorFlow, PyTorch, scikit-learn, or similar ML frameworks.
• Clear written communication that makes complex ML concepts precise and understandable.
• Careful attention to correctness and comfort working independently in an asynchronous environment.

Helpful background

Three to five or more years of work on ML projects, pipelines, or MLOps is preferred. Experience with model evaluation, cloud deployment, production ML systems, code review, or technical writing is also useful.

A bachelor's, master's, or PhD in computer science, machine learning, or a related field is valued. Data labeling, RLHF, and other AI training experience can also be relevant.

Contract outlook

The work combines freelance autonomy with defined technical tasks for AI research and enterprise applications. Strong contributors may take on broader responsibilities and lead new programs, while additional projects may offer ongoing work or contract extensions.