This freelance role involves building realistic, structured test scenarios that check how LLM-based agents perform tasks a human would normally do. The work also covers defining the ideal path and acceptable behavior, then reviewing agent outputs and refining the tests. It suits students or graduates in computing, data or AI who are comfortable with analytical, detail-focused work and can work remotely on flexible hours.
Pay not disclosed
The platform does not publish the pay for this job.
Open to residents of: Canada, United Kingdom, Greece, Ireland, Italy, Singapore, United States.
You apply on Mindrift, the AI training platform run by Toloka: submit your application for the role, pass a qualification step, then work on projects that match your expertise.
Official posting: https://apply.workable.com/j/75513B02E3
Last checked on October 16, 2026.
Review and evaluate AI-generated outputs across text, images, and structured data to assess quality, accuracy, and consistency. This remote contract role is ideal for detail-oriented individuals who can identify errors and provide actionable feedback to guide AI model improvement, with no prior AI experience required.
This role involves completing structured evaluation and labeling tasks to train and improve AI systems. You will assess AI-generated content across multiple formats and provide feedback to help make AI systems more accurate and reliable, working remotely on a flexible schedule.
This role involves evaluating how well AI models can perform computational biology tasks. Computational biologists and bioinformaticians will design realistic analysis scenarios from their own work, run them through AI systems, and grade the results against professional standards to help benchmark model capabilities.
This contract role asks QA specialists to test AI system outputs against defined requirements, label and review datasets, and report model errors and patterns. It is aimed at people with software quality assurance and domain knowledge, and no prior AI experience is needed.
Posted October 16, 2026
Referral link to micro1's job list: search for this role there. Open this exact job
This freelance role involves designing structured test scenarios that simulate real-world tasks performed by humans, so that LLM-based agents can be evaluated against defined gold-standard behavior. The work also covers annotating steps, expected outputs and edge cases, and refining tests alongside developers. It suits students and early-career candidates in computer science, data science or related fields who have strong analytical skills and advanced English.
Posted October 15, 2026
This freelance role involves designing structured evaluation scenarios that test how LLM-based agents carry out tasks a person would normally perform. The work includes defining the ideal path, acceptable agent behavior and edge cases, then refining them with developers. It suits candidates with a computer science, AI or data background who are comfortable with complex guidelines.
Posted October 15, 2026
