About this opportunity
Description
About the Role
Sovrano is building a pool of qualified annotators for upcoming robotics data projects supporting AI companies developing robot learning models.
The work involves reviewing short video clips and accurately describing actions performed by robots or human hands, either frame by frame or through captions. Precision, consistency, and the ability to follow detailed annotation guidelines are important.
Please note: This is a pre-qualification opportunity. There is no immediate project guaranteed. Successful applicants will join Sovrano's qualified expert pool and may be contacted when relevant projects become available.
What You'll Do
- Describe robotic and manual actions shown in short video clips.
- Annotate actions frame by frame or using captions.
- Follow project labeling guidelines consistently.
- Maintain accuracy throughout annotation sessions.
- Flag unclear, corrupted, or low-quality clips rather than guessing.
- Meet agreed turnaround times when assigned to a project.
- Raise questions when examples do not clearly fit the guidelines.
Qualification Process
Applicants must complete a free, unpaid qualification on Sovrano's partner annotation platform. This includes:
- English assessment.
- Typing assessment.
- Optional practice tasks.
- A 5-task qualification exam.
You must register on the partner platform using the same email address used for your Sovrano application so your results can be verified.
After completing the exam, you'll also need to submit the completion form provided in your application email, including the required account information and proof of exam completion.
Requirements
- Working written and reading English.
- Strong attention to detail.
- Ability to follow written annotation guidelines precisely.
- Laptop or desktop computer with reliable internet.
- Ability to work independently and remotely.
- Ability to set up two-factor authentication on your phone.
- Availability for at least a few hours per week when projects become available.
Previous experience with data annotation, video annotation, data QA, robotics, or computer vision datasets is helpful but not required.
Contract & Compensation
This is a freelance, project-based and fully remote opportunity.
The qualification process is unpaid. If you're subsequently allocated to a project, compensation ranges from $5–$15 per hour, depending on experience and assessed accuracy. Your rate will be confirmed before work begins.
Hours vary by project and are agreed before starting. Payment is made monthly for approved hours.
Important: Completing the qualification does not guarantee a project, offer of work, or specific number of hours.
