The Work
You will review financial documents and check that extracted text, labels, and structure are accurate, complete, and meaningful. The work covers invoices, receipts, bank statements, financial statements, tax and audit materials, payment records, and business reports.
- Correct OCR errors in numbers, dates, currency amounts, account names, table values, totals, and financial terms.
- Assign labels to document regions such as text, headers, tables, key-value fields, footnotes, signatures, handwritten text, and unreadable content.
- Review or create clean HTML tables that preserve rows, columns, headers, totals, and financial meaning.
- Create JSON key-value pairs for invoice numbers, dates, vendor and customer names, totals, balances, payment terms, tax amounts, and account references.
- Organize reading order and parent, child, or previous-instance relationships so documents can be reconstructed correctly.
- Complete a final quality check with close attention to financial accuracy and consistency.
What It Pays and Takes
Accounting and finance knowledge matters more than experience with a particular annotation tool. The tool provides OCR preannotation and bounding boxes, but you will verify, correct, and complete the output yourself.
- Pay: $7 to $9 USD per hour.
- Time: 20 or more hours per week.
- Work type: Part-time contractor.
- Experience: Intermediate level.
- Language: English fluency.
- Location: Open to candidates in the countries specified for this role.
- Required knowledge: Accounting and finance, including financial statements, invoices, receipts, banking and payment records, bookkeeping, audit, tax, OCR, tables, and key-value document structure.
How It Works
Apply on OpenTrain. The employer reviews applications there.
About AI Training Work
AI training work uses examples reviewed and prepared by people to help artificial intelligence systems understand documents and produce useful results. Accounting and finance contributors are paid for applying careful domain knowledge to check financial details, document structure, and model outputs.