How Studio Works

Scale Studio helps you unlock your labeling team’s efficiency with the most comprehensive labeling platform. If you want to label data yourself, or with your own team, Studio is a great solution.

--

Bring your own data
Upload any data set you need labeled in order to progress your AI/ML project. Studio supports a wide range of project types, including text, image, and video. You can view more details in the API documentation.

--

Bring your own labelers
Onboard your own team of labelers onto the Studio platform to label your data. Whether they are your own internal labeling team, or an outsourced workforce, you can invite them onto your project and assign them to work as attempters or reviewers.

• Attempters follow the instructions to label your data. The project you create will primarily be focused around structuring how the attempters label your data.
• Reviewers review labeling attempts. Reviewers do a final check of the attempters’ work and maintain the overall quality level of the project.

--

Monitor project and labeler performance
On Scale Studio, administrators can:

• See how individual labelers have performed.
• Monitor metrics such as total submissions, accuracy, and submission time at a project-level or on a per-labeler basis.
• Easily assign your labelers different roles based on their performance
  Labelers can also see their task history and review their performance on your project (e.g. they can see their overall accuracy scores (via quality tasks), and track their improvement over time.) Attempters can also see the feedback that reviewers have left for them in order to continuously improve.