Review and rate AI-generated Portuguese responses, write clear model solutions, and apply localization expertise to ensure accuracy and cultural nuance. Part-time contractor role, fully remote, based in Brazil; up to $30/hour and 20+ hours per week.
Generative AI & RLHF
Remote Hourly · $30/hr
$30/hr
Compensation
2 countries
Eligibility
Intermediate
Experience
Jul 9, 2026
Posted
Open to applicants in
Brazil Portugal
Interested in this role?
Create a free OpenTrain account and apply in minutes.
OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We help people start and grow freelance careers teaching AI by centralizing projects, building a unified profile, and making specialized work easier to track and apply for.
About AI training work
AI training (also called data labeling or human feedback work) is the human side of building modern AI. Contributors annotate, evaluate, and improve model outputs across text, image, audio and other data; this role focuses on evaluating and improving generative text and reasoning in Portuguese.
Flexible, remote work that fits around other commitments
Directly shape how AI systems behave by providing judgment, edits, and model solutions
The role
We are recruiting a Portuguese AI Response Evaluation Expert to evaluate AI-generated responses for correctness, logical reasoning, clarity, and adherence to prompts. You will compare responses, identify methodological or factual errors, write high-quality explanations and model solutions, and apply Brazilian Portuguese localization judgment.
Work with RLHF and evaluation-rating tasks for text generation
Focus on preserving correct terminology, tone, and cultural nuance in Brazilian Portuguese
What you'll do
Daily work centers on evaluating AI answers, fact-checking localized details, and producing explanatory feedback and model solutions that help improve model behavior.
Review AI-generated responses and language outputs for accuracy, logic, and clarity
Evaluate step-by-step problem solving, methodology, and conceptual understanding
Compare multiple AI responses and rank them by correctness and reasoning quality
Fact-check names, dates, units, references, and other localized content
Write high-quality explanations, corrections, and model solutions when required
Apply Portuguese localization and translation judgment to preserve terminology, tone, and nuance
Requirements
Candidates must meet the listed skills and experience exactly as part of evaluation and quality assurance workflows.
Degree in Translation, Linguistics, Localization, Communications, or a related field
Native or near-native Portuguese proficiency (Brazilian Portuguese focus)
C1 English reading and writing proficiency for prompts and evaluation guidelines
5+ years of localization or translation experience
MQM/LQA knowledge including severity, category, and root-cause analysis
Experience with CAT tools, translation memory, termbases, and QA workflows
Excellent editorial judgment and strong fact-checking skills
Experience with AI data training, annotation, or evaluation is preferred
Helpful background & contractor expectations
Ideal applicants have hands-on experience with terminology management, translation memory, style guides, automated checks, and copyediting. We expect responsive, well-documented, and dependable hourly contractor work habits.
Familiarity with termbases, translation memory and automated QA checks
Clear documentation of decisions and root-cause analysis
Dependable contractor mindset: responsiveness, logging hours, and clear communication
Logistics, pay, and how to apply
This is a part-time hourly contractor role. The project requires 20+ hours per week and is fully remote. The role is based in Brazil and candidates in Brazil or Portugal may apply; worldwide applicants are not accepted.
Compensation is hourly at up to $30.00 USD per hour. You will work on evaluation tasks focused on RLHF, evaluation-rating, and text generation labels and follow provided guidelines and QA workflows.
Employment type: Contractor, Part-time
Time requirement: 20+ hours/week
Location: Fully remote; role based in Brazil (candidates in Brazil and Portugal eligible)
Join OpenTrain as a part-time AI Quality Analyst evaluating a conversational AI personalization feature; Portuguese fluency required. Remote, contractor work at $15/hr with a 20+ hour weekly commitment—you'll design personal prompts, compare side-by-side responses, and write defensible rationales.
Join OpenTrain as a part-time, contract Portuguese QA Annotator reviewing Portuguese AI outputs for quality, correctness, and guideline alignment. Flexible hours (<20 hrs/week), $15/hr, must reside in Portugal with C1+ European Portuguese and B2+ English.
Lead QA for Portuguese AI training projects: review AI-generated Portuguese content and trainer work, give precise written feedback, maintain style guides, and improve QA workflows. $35/hr, 20+ hours/week, remote (Brazil or Portugal), contractor, part-time.
Brazil 
Portugal