* Salário: R$ 2.000 a R$ 5.000 por mês (estimado)
* O valor exibido é uma estimativa calculada com base em dados públicos e referências do mercado. Não garantimos que este seja o salário oferecido para esta vaga específica.
Área: Outros
Nível: Senior
About the Role:
TrustScale is building scalable, high-quality human-in-the-loop systems to power AI. We are looking for a Brazilian Portuguese Quality Control Specialist who can analyze, validate, and challenge AI-generated outputs with precision and consistency. This is not a proofreading role. You will act as a quality gatekeeper, ensuring outputs meet standards for accuracy, utility, and linguistic quality across large-scale workflows.
What You Will Do:
- Evaluate AI-generated content in Brazilian Portuguese (PT-BR) across multiple task types (annotation, review, validation).
- Apply structured quality frameworks to assess:
- Accuracy and factual correctness.
- Utility (alignment with user intent).
- Language quality (fluency, tone, clarity).
Identify and flag:
- Hallucinations and misinformation.
- Logical inconsistencies.
- Cultural or linguistic mismatches.
- Provide clear, structured feedback to improve upstream quality.
- Detect patterns of errors across batches and contribute to quality insights.
- Support calibration efforts to ensure consistent scoring across teams.
- Contribute to guideline refinement and evaluation standards.
Who You Are:
- Native Brazilian Portuguese (PT-BR) with strong command of English.
- Highly analytical with strong critical thinking skills.
- Comfortable working with ambiguous and evolving guidelines.
- Detail-oriented with the ability to maintain consistency at scale.
- Clear communicator, able to justify decisions and provide actionable feedback.
- Naturally skeptical—able to question outputs and validate information.
Preferred Background:
- Journalism, editorial, or investigative research.
- Linguistics, translation, or localization QA.
- Content moderation / Trust & Safety.
- Experience in AI data annotation, evaluation, or QA.
What Success Looks Like:
- High consistency in quality scoring across tasks.
- Strong alignment with QA benchmarks.
- Ability to detect non-obvious quality issues.
- Feedback that improves annotator performance and overall output quality.
- Contribution to a scalable, predictable quality system.
Why TrustScale:
At TrustScale, we operate AI talent as a supply chain—structured, measurable, and optimized. You will be part of a system where quality is not subjective, but defined, measured, and continuously improved.