Create audit questions and case studies that test how AI models handle real audit work, from planning and risk assessment to controls, evidence, and fraud detection. You will assess model answers against GAAS and audit best practices, then help audit and technical specialists shape the difficulty and learning value of each challenge.
Mercor keeps this listing open as a standing pool for language and audio specialists, rather than for one active project. After you submit a resume, confirm where you work from, and complete an AI interview of about 20 minutes, Mercor verifies your background and may invite you to a project that matches your language or specialty. Project matches can come within a week or take several months.
Review and compare AI-generated answers, check their reasoning and factual accuracy, and create Portuguese-language examples that show correct methods. Your Portuguese localization expertise will guide decisions on terminology, tone, and cultural context, while MQM and LQA methods help you assess errors consistently. This remote contract role is open only to candidates based in Brazil; there is no immediate project, but qualified experts may be contacted for future work.
Use your music or audio experience to document what you hear in 30–90 second field recordings, from instruments and musical details to sounds in the surrounding scene. You’ll enter annotations in a project web platform, compare recordings of the same song from different locations, and later assess AI-generated descriptions. The contractor engagement calls for 20–25 hours per week.
Create and assess accounting challenges that test how AI models apply GAAP to financial statements and regulatory disclosures. You will draw on financial reporting expertise across topics such as revenue recognition, lease accounting, consolidations, fair value, and SEC requirements. The role is remote and asynchronous, with flexible scheduling.
Reviewers listen to Dutch (Netherlands) AI-narrated audiobooks and assess whether the speech sounds accurate, natural, and enjoyable. You’ll mark and categorize speech errors, then share where narration holds a listener’s attention or disrupts the experience. The remote, part-time schedule calls for about 20 hours per week, or roughly four hours per day.
Review AI-generated Odia responses and create reference answers for tasks such as customer support, education, and general knowledge. You will judge grammar, meaning, tone, cultural fit, and factual reliability, then explain issues or provide improved Odia wording in English. The role calls for native or near-native written Odia, strong English, and at least three years of relevant language-focused experience.
Review Vietnamese audio for speaker or model fluency, pronunciation, intonation, and linguistic authenticity. You’ll record clear English feedback with reasons for each assessment, follow the project’s guidelines, and share language insights that support Vietnamese AI training. No previous AI experience is needed.
Create and solve data science problems based on practical work in experiments, metrics, forecasting, segmentation, and predictive modeling. You will provide a reference solution with reasoning for each step and identify plausible analysis errors that require hands-on practitioner experience. The role suits practicing data scientists, not data engineers, ML engineers, or reporting-only analysts, and requires at least 10 hours of work per week.
Join a standing pool of practicing lawyers for AI training and evaluation projects, rather than applying to one active opening. You may draft legal documents and grading references, assess model-written work, or explain why a legally accurate answer still fails in context. After you submit a resume, confirm your work location, and complete a roughly 20-minute AI interview, Mercor verifies your background and license before considering you for matching projects. Project invitations may come within a week or take several months.
Use your Punjabi language expertise and mathematics education to shape AI training materials and assess model responses. You will write prompts and reference answers, compare generated content, and give precise feedback on grammar, meaning, reasoning, tone, and cultural fit. The work requires native or near-native Punjabi, English proficiency at C1 or higher, and at least three years of relevant professional experience.
Review audio samples of AI-generated Indonesian speech and judge how natural, fluent, and culturally appropriate they sound. You’ll give written and verbal feedback on language accuracy, pronunciation, intonation, and authenticity, then record clear observations to guide model improvements. The work supports a large-scale effort to train AI systems; prior AI experience is not required.
Create marketing questions and case studies for AI training, then assess how well models apply marketing principles to realistic business scenarios. You will cover areas such as brand strategy, digital marketing, consumer behavior, and campaign optimization, and give detailed feedback on the quality of model responses. The role calls for hands-on marketing expertise and clear written communication.
Mercor’s Marketing Specialist network connects experienced marketing professionals with future contract projects at AI labs and companies. After you complete your profile and pass an AI interview, Mercor may match you with projects as they arise. Work can cover AI model training and evaluation, marketing deliverables, and expert feedback. Project scope and duration vary, with typical commitments of 15 to 30 hours per week.
Create realistic Blender assignments that test professional 3D workflows, then record your own work with the provided capture tools. You will document reproducible steps and define clear checks for meshes, scenes, and renders, helping train and evaluate AI agents on desktop software. The contract requires a Blender-capable computer running Windows 11, Ubuntu (X11), or macOS, plus B2 English proficiency.
Review Portuguese audio for a project that supports AI language understanding and generation. You’ll judge how natural and fluent the speech sounds, assess pronunciation and intonation against set guidelines, and explain your findings in English. Prior AI experience is not required; the work draws on your knowledge of Portuguese.
AfterQuery needs construction professionals with 1 to 3 years of experience for project-based work over the next 2 to 3 weeks. You’ll review construction documents, support planning and field problem-solving, and evaluate technical questions across different project types. Work runs asynchronously, with a 10 to 20 hour weekly commitment for each project you accept.
Mercor seeks Economics Research Experts for a project with a leading foundational model AI lab. Candidates need a PhD in Economics and at least two years of experience at a top institution leading economics research, such as the World Bank, IMF, Federal Reserve, or a graduate school. The engagement lasts at least four weeks and requires a commitment of 10 or more hours per week.
Create reference answers in Assamese and assess AI-generated responses across general knowledge, customer support, education, and conversational writing. You will judge meaning, grammar, tone, cultural fit, and prompt adherence, then explain issues in English. The role requires native or near-native Assamese, C1 English, and at least three years of relevant language work.
Review Korean audio for a project that helps train AI systems. You’ll judge how natural and fluent a speaker or model sounds, assess pronunciation and intonation, then explain your findings in English. The work follows established guidelines and draws on Korean language expertise; prior AI experience is not required.
Location and eligibility
Remote · Eligibility not specified
Engagement type
Contract
Compensation
$30–$65/hr
Korean
Language & LinguisticsView job
Showing 20 active roles of 801. Dates reflect what the source published, not when AIWorkJobs last synced.
Sources: ten specialist AI work platforms. AIWorkJobs does not verify every claim made by a source. Report a listing that looks wrong through the contact page.