生成式AI分析师 - 土耳其语（土耳其）
Generative AI Analyst - Turkish (Turkey)
我们正在寻找生成式AI分析人员，以支持跨多媒体数据（例如视频、图像和相关元数据）的标注和质量审查。该职位最适合具有多年标注和质量保证经验、注重细节且能在高工作量下熟练使用结构化指南的候选人。
主要职责
- 使用定义的标注规则准确标注和审查多媒体数据（例如事件标签、时间边界、对象/实体属性、一致性检查）。
- 一致遵循详细指南；对模糊案例和边缘情况进行标记，并提供清晰示例以供解决。
- 对已完成的工作进行自我质量检查，并在提交前纠正错误。
- 参与校准会议，统一指南解读，减少评审者之间的差异。
- 跟踪重复性问题，并通过记录示例、故障模式和建议的澄清说明，为指南改进做出贡献。
- 支持质量工作流程（抽样、二次审查、审计），并将反馈融入后续工作中。
- 在需要时与内部利益相关者和外部合作伙伴协调，以满足吞吐量和质量目标。
项目详情
持续时间：持续进行
用工类型：自由职业
地点：远程
语言：土耳其语
申请后，您将成为我们社区的一员，开启一系列针对您技能和可用性的项目。
加入我们意味着为当前项目做出贡献，并成为我们充满活力的网络的一部分。
这是一个独特的机会，可以提升全球用户体验，并以有意义的方式应用您的语言技能。
资格要求
- 具有多媒体标注、多模态数据标注、计算机视觉标注、内容质量保证或相关领域的3年以上经验。
- 在重复性、细节导向的工作中表现出保持高准确性和一致性的能力。
- 在土耳其语方面具备良好的书面和口头沟通能力（能够清晰地记录问题并提出精确的问题）。
- 能够长时间处理多媒体内容，同时保持专注和准确性。
- 有按照结构化指南工作并满足生产率/质量期望的经验。
最初发布于Himalayas
查看英文原文
OVERVIEW
We are looking for Generative AI Analysts to support labeling and quality review across multimedia data (e.g., video, images, and related metadata). This role is best suited for candidates with several years of annotation and QA experience, strong attention to detail, and comfort working with structured guidelines at high volume.
MAIN DUTIES
- Annotate and review multimedia data accurately using defined labeling rules (e.g., event tagging, temporal boundaries, object/entity attributes, consistency checks).
- Follow detailed guidelines consistently; flag ambiguous cases and edge scenarios with clear examples for resolution.
- Perform self-QA on completed work and correct errors before submission.
- Participate in calibration sessions to align interpretation of guidelines and reduce reviewer-to-reviewer variance.
- Track recurring issues and contribute to guideline improvements by documenting examples, failure modes, and recommended clarifications.
- Support quality workflows (sampling, second-pass review, audits) and incorporate feedback into subsequent work.
- Coordinate with internal stakeholders and external partners as needed to meet throughput and quality targets.
Project Details
Duration: Ongoing
Employment Type: Freelance
Location: Remote
Language: Turkish
By applying, you'll become part of our community, opening doors to a range of projects tailored to your skills and availability.
Joining us means contributing to our current project and becoming part of our dynamic network.
This is a unique chance to enhance global user experiences and apply your language skills in meaningful ways.
Qualifications
- 3+ years of experience in multimedia annotation, multimodal data labeling, computer vision labeling, content QA, or a closely related field.
- Demonstrated ability to maintain high accuracy and consistency across repetitive, detail-oriented work.
- Strong written and verbal communication in Turkish (to document issues clearly and ask precise questions).
- Comfortable working with multimedia content for extended periods while maintaining focus and accuracy.
- Experience working from structured guidelines and meeting productivity/quality expectations.
Originally posted on Himalayas