Paro

Accounting - AI Trainer

US$83,200-$260,000Posted 26 days ago

Remote Temporary AI Training & Annotation $200k+

Apply for this role

Job Description

Accounting AI Trainer

*Contract · Remote · $40–$125/hr USD*

About the role

We are seeking experienced accounting professionals for a flexible AI training opportunity focused on evaluating how frontier AI models handle real-world accounting work, including technical accounting conclusions, close and reconciliation, audit procedures, tax positions and provisions, and financial reporting.

You’ll bring the professional judgment you’ve developed through real accounting work—clearing review notes, tying out workpapers, analyzing financial data, and defending conclusions to auditors, partners, leadership, or audit committees—and use that expertise to evaluate AI-generated work.

In this role, you may design challenging, realistic tasks drawn from your own professional experience, such as a month-end close reconciliation package, technical accounting memo, tax provision workpaper, audit sampling plan, disclosure checklist, or management letter comment. You’ll then evaluate how frontier AI models perform against the professional standards you would expect from a qualified colleague.

You will work with realistic professional files and scenarios similar to those practitioners encounter in their day-to-day work. Some assignments may be compact and involve only a handful of files, while others may involve larger scenarios completed over multiple days. The goal is to identify areas where sophisticated AI models still struggle with work that an experienced accounting professional should be able to complete correctly.

This is not a traditional accounting role. You will be helping build better AI by putting your knowledge to work in a structured, flexible, fully remote environment. The work is long form and self directed, and clear written reasoning matters as much as technical depth.

Responsibilities

Design challenging, realistic accounting tasks drawn from your own day to day work: the scenario, a prompt phrased the way you would brief a trusted colleague, and the supporting files a professional would need (trial balances, schedules, source documents, prior year workpapers, correspondence), which you author yourself.
Run those tasks through frontier AI models and evaluate the deliverable they produce (the workpaper, memo, schedule or reconciliation) against the standard you would hold a colleague to.
Compare two model outputs on identical prompts and files, decide which performed better, and document where each fell short.
Write detailed grading rubrics that specify what a correct deliverable must contain, such as the right treatment applied, the right amounts, the right disclosures and the right support, and explain in writing why a response passes or fails each one.
Flag concrete failures with evidence: misapplied standards, numbers that do not tie, unsupported conclusions, fabricated or ignored source documents, missed disclosures, and off brief interpretation of the ask.
Contribute across audit, technical accounting, tax, close and reporting, and review and refine tasks built by other experts.

Domain qualifications

2+ years of hands on experience preferred in public accounting, corporate accounting, audit, tax, or financial reporting.
In progress Bachelor’s degree or higher in accounting, finance or a related field.
Depth in at least one of: audit and assurance; technical accounting and financial reporting (US GAAP and/or IFRS); tax (individual, corporate, partnership, provision); controllership, close and consolidation; cost and managerial accounting; internal audit and SOX; forensic accounting.
Working understanding of several of the others, enough to know what those workflows involve and how they are run, so you can assess work in an adjacent area and point out what was done correctly or incorrectly.
Comfortable in spreadsheets at a practitioner level: building and checking schedules, reconciliations and roll forwards, not only reading them.
CPA, CA, ACCA or equivalent is preferred but not required. Practical expertise outweighs credentials. Credential holders may be asked to verify the credential during the assessment.

General requirements

2+ years of hands on experience in your field preferred (see Domain qualifications above). Candidates with less experience are considered where the practical work is real.
Able to draw on your own real world experience and day to day workflows to craft scenarios that test whether an AI system can actually do the work.
Hands on practitioner: you currently do (or recently did) the work yourself at an individual contributor level, not solely in a managerial capacity.
Full professional or native level written and spoken English, with strong written communication. You can explain complex professional reasoning clearly and concisely, and articulate why a result is wrong, not only that it is.
Comfort with ambiguity and attention to detail. You can orient in a new set of files and build an accurate, deep working picture of it quickly, especially when the subject sits partly outside your own specialization. You verify what a document claims against the underlying numbers, sources or facts.
Capable of interpreting feedback, judging which parts of it are actually correct, and applying it without hand holding. When stuck, you look for the answer rather than waiting for one.
Ability to ramp quickly on unfamiliar work from written material and instructions alone, including where that material is incomplete (for example, writing grading rubrics for the first time).
General familiarity with AI and LLM tools. You have used models like Claude or ChatGPT in professional work and have the judgment to tell a well reasoned answer from a plausible sounding but incorrect one.
Baseline tech literacy: comfortable with cloud file tools (e.g., Google Workspace), managing browser profiles, downloading and installing desktop apps (e.g., Claude), and everyday file handling (e.g., converting between Excel and Google Sheets, zipping files for sharing).
Available at least 10 hours per week, with no weekly maximum. Consistent availability is valued and full time hours are available.
Based in the United States, Canada, or the UK.

Compensation and terms

Pay: $40-$125/hr USD. Paid via PayPal on a regular cadence.
Contract position. Fully remote. Minimum 10 hours per week with no weekly maximum.
Flexible scheduling. Multi day task timers let you spread work across days.
Access to Claude and ChatGPT is provided through the project; no personal subscription is required.

What to expect

Apply through your agency link and create your account.
Complete the skills assessment (about 2 to 4 hours, unpaid). It tests domain fit, careful reading, and how you judge a model's output against a professional standard. All work must be your own; submissions produced with AI tools are rejected, and this is the single most common reason candidates do not pass.
Our team reviews your assessment. If you pass, you complete onboarding, start on regular paid projects in your field, and complete a short training project that walks through how tasks, files, and rubrics are built.
Propose a task from your professional experience: the scenario, the prompt, the files you will build, and where you expect the model to fail. An expert reviewer reads it and sends written feedback either way.
Once your proposal is accepted, build the task in full and run it against frontier models. Every completed task goes through expert review, and experienced contributors are invited to review and refine other experts' tasks.
Each stage is a gate: work must be accepted before you move on. You are compensated for every step you complete, but not every submission is accepted. Feedback and revision are a normal part of the process. Support is available through platform instructions, onboarding materials, a dedicated Slack channel, and office hours.

Pay: $40.00 - $125.00 per hour