Data annotation job
Accounting Expert
Mercor · Posted Yesterday
$95/hrRemote (US)Contract
- AI evaluation
- LLM training
- Vision
- Finance
About this role
Role Overview
- Mercor is hiring experienced accountants and auditors to build and review evaluation tasks for AI agents.
- Each task is set in a realistic simulated company, with the ledgers, workpapers, policies, emails and business apps an accounting team really uses.
- You make sure each task reads like a real request, the reference answer holds up against the source records, and the grading rubric checks what a senior accountant would check.
Key Responsibilities
- Refine AI-drafted task prompts so they read like a real request from a controller, CFO or audit lead, without giving the answer away.
- Check the reference answer against the world's files and app records, and correct what is wrong.
- Write and tighten grading rubrics: pass/fail criteria with exact values, plus the judgment calls a generalist would miss.
- Read how leading AI models attempt each task and confirm they fail for fair, defensible reasons.
- Work across financial reporting and technical accounting (revenue recognition, leases, consolidation, business combinations, SEC reporting), month-end and year-end close, internal controls and SOX, tax provisions, audit fieldwork, and industry areas such as fund, nonprofit, construction and SaaS accounting.
Ideal Qualifications
- 3+ years as an accountant, controller or audit partner. Big Four or major-corporation experience preferred.
- Bachelor's degree in accounting, finance or a related field.
- Active CPA license, or an equivalent such as CPA Canada, CPA Australia, CA, ACCA or ICAEW.
- Hands-on ownership of financial reporting, audit engagements or technical accounting, with working knowledge of US GAAP / IFRS, SEC filings and PCAOB / ISA standards.
- Prior experience on a Mercor RL project is a plus.
Details
- Pay: $95/hour.
- You start with a short onboarding session and a paid skills assessment. Once you pass, you can claim your first task right away.
- Strong performers can move into reviewer and world lead roles.
This listing is aggregated by DataAnnotationJobs.org. Applications are processed by Mercor, not by this site.
Mercor is an AI-native talent marketplace connecting vetted domain experts such as doctors, lawyers, and engineers directly with frontier AI labs for RLHF, evaluation, and specialized training data at massive scale. See all Mercor jobs.
Apply for this role
Your details first, then a short screening test. Both are read as part of your application and go to Mercor, and we will tell you when similar roles appear.