Opentrain AI
Design realistic data entry and validation tasks that test AI accuracy, completeness, and data integrity. This remote contractor role pays $20 to $35 per hour and requires strong English, careful review skills, and weekly availability.
The work
You will create benchmark tasks that train and assess AI agents on realistic data entry and validation problems. Your work will turn complex accuracy and reconciliation challenges into structured tasks with clear expected results.
You will work with files such as CSVs, PDFs, spreadsheets, and technical documents. Tasks should include realistic issues such as malformed records, missing information, inconsistent formats, silent truncations, and other defects found in data-sensitive work.
- Design expert-level data entry and validation tasks.
- Define the correct final state for each task.
- Identify the errors an AI agent should find and the reconciliation steps it should complete.
- Write detailed grading rubrics with at least 35 criteria focused on accuracy and completeness.
- Make evaluation materials reflect compliance-sensitive and audit-sensitive work.
- Refine tasks and evaluation materials through precise written and verbal communication in an asynchronous setting.
What it pays and takes
This is a part-time contractor role with output-based compensation. The structured rate is $20 to $35 USD per hour, while the role description states that payment is tied to tasks that meet project specifications.
- Pay: $20 to $35 USD per hour in the structured listing; payment is tied to accepted tasks meeting project specifications.
- Availability: The description sets a minimum of 10 to 15 hours per week, while the listing also states 20+ hours per week.
- Location: Remote, in the countries listed for this role.
- Language: Proficient written English is required.
- Experience: Extensive experience in data entry, quality assurance, or data validation.
- Skills: Detect inconsistencies, formatting problems, missing data, and silent truncations across file types.
- Requirements: Demonstrated ability to achieve and document high accuracy standards, including measurable error rates and outcomes.
- Work style: Exceptional attention to detail, a disciplined process, independent work, and proactive communication.
- Helpful background: Healthcare claims, finance back-office operations, legal operations, compliance-sensitive work, audit-sensitive processes, or other high-stakes data integrity work.
- AI training experience is not required. Practical domain knowledge and a strong record of accurate validation are valuable.
- Selected contributors are expected to begin their first tasks shortly after onboarding.
How it works
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI training work
AI training is the human work behind systems that learn from examples, including checking whether an AI response or action is accurate and complete. People with strong data validation and specialist knowledge help set reliable standards for these systems.