Data Entry AI Evaluation Task Designer
The Work
You will create benchmark tasks that show whether AI agents can complete and validate data-entry work accurately. Your tasks will use realistic defects and define the correct final result, required reconciliation steps, and the errors an agent must find.
You will also create detailed grading materials and refine them through clear written and verbal communication in a distributed work setting.
- Design expert-level tasks based on real-world data-entry and validation challenges.
- Build datasets from CSV files, PDFs, spreadsheets, and technical documents.
- Include malformed records, missing information, inconsistent formats, silent truncations, and other realistic defects.
- Define the correct final state and the errors and reconciliation steps an AI agent must identify.
- Write grading rubrics with at least 35 criteria focused on accuracy and completeness.
- Make sure evaluation materials reflect compliance-sensitive and audit-sensitive work.
What It Pays and Takes
This is a part-time contractor role. The role is listed as entry level, but the work requires extensive practical experience in accurate data entry, quality assurance, or data validation. Previous AI training experience is not required.
- Pay: $20 to $35 per hour.
- Availability: 10 to 15 hours per week, with a minimum weekly submission requirement.
- Work arrangement: Part-time contractor work with output tied to tasks that meet project specifications.
- Language: Proficient written English for drafting complex tasks and comprehensive rubrics.
- Required: Experience detecting missing data, formatting problems, inconsistencies, and silent truncations across file types.
- Required: A record of maintaining and documenting high accuracy, including measurable error rates and outcomes.
- Required: Exceptional attention to detail, a disciplined process, independent work habits, and proactive communication.
- Helpful experience: Healthcare claims, finance back-office operations, legal operations, compliance-sensitive work, audit-sensitive processes, or other high-stakes data integrity environments.
- Location: Open to candidates in AE, AR, AT, BD, BE, BH, BR, CA, CH, CL, CO, DE, DK, DZ, EG, ES, ET, FR, GB, GH, GR, ID, IN, IT, JO, JP, KE, KG, KW, KZ, LB, LK, MA, MX, MY, NG, NL, OM, PE, PH, PK, PS, PT, QA, SA, SG, TH, TN, TW, US, UZ, VN, and ZA.
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 reviewing model results and testing whether an AI follows instructions accurately. OpenTrain AI is the hiring and contracting organization for this role, and people with strong subject knowledge and careful review skills help define the standards AI systems must meet.