About The Company Jobrapido connects people with employment opportunities and supports organizations in reaching talent. For this flexible contract project, Jobrapido is seeking experienced machine learning researchers to contribute specialized expertise to the evaluation of frontier artificial intelligence. The work focuses on assessing whether advanced AI models can perform authentic research tasks in a command-line environment. Researchers will develop rigorous, realistic evaluations that help reveal how models handle the reasoning, implementation, and analysis involved in machine learning research. About The Role As a Senior Machine Learning Researcher, you will create research-grade tasks designed to test the limits of a leading AI model. Your work will draw on practical research experience, including interpreting technical methods, implementing algorithms, diagnosing experimental problems, and assessing empirical results. You will build complete, self-contained task environments, run the tasks against a frontier model, and examine its attempts with the care expected in scholarly peer review. This is a remote, flexible contract opportunity with an expected commitment of 10 to 40 hours per week. You will collaborate with reviewers and refine each task through the acceptance process. The objective is to produce evaluations that are challenging for meaningful reasons: a model’s failure should demonstrate a genuine capability gap, not unclear instructions, missing context, or misleading wording.

Qualifications

PhD in machine learning, computer science, statistics, or a closely related discipline. Academic or industry experience conducting machine learning research; doctoral research experience counts toward this requirement. At least one research publication, preferably peer-reviewed work in a recognized machine learning venue or journal. Strong Python skills and substantial experience with at least one major machine learning framework, such as PyTorch or JAX. Fluency with Linux command-line tools, shell scripting, Docker, and Git. Ability to assess technical work critically, communicate findings clearly, and develop precise, reproducible task specifications. Experience as a peer reviewer or in designing course assignments and automated grading systems is a strong advantage.

Responsibilities

Design self-contained command-line tasks based on authentic machine learning research practices. Examples include implementing a method using only a paper’s description, diagnosing a flaw in an experimental setup, reproducing a reported result and explaining discrepancies, implementing a custom loss function or optimizer, and conducting a statistically sound comparison between models. Build each task environment by preparing the code, data, and dependencies needed to complete it. Write a reference solution and automated tests that reliably determine whether a proposed solution is correct. Make tests deterministic where possible, and use well-justified tolerances when results naturally vary. Run tasks against a frontier AI model, inspect its approach and output, and document the reasons for success or failure. Refine task wording, implementation, and evaluation criteria to ensure the challenge is fair, clear, technically valid, and genuinely demanding. Work with reviewers to address feedback and bring each task to acceptance.

Benefits

Contribute research expertise to evaluations that provide insight into the capabilities of frontier AI models. Expand your professional influence through evaluation projects, advisory opportunities, and research collaborations. Work alongside researchers from leading institutions and laboratories on technically demanding projects. Set a schedule that fits around your research position and other commitments, with flexible work ranging from 10 to 40 hours per week. Compensation is competitive, with rates determined by project scope, skills, and experience. Equal Opportunity Jobrapido is committed to a fair and respectful contracting process. Applicants will be considered based on relevant qualifications, experience, and project requirements. We welcome qualified candidates from diverse backgrounds and are dedicated to providing an inclusive professional environment.