Domain Expert, AI Evaluation Task Writer: Finance (Contract, Remote)

Patronus AI, Inc.

5d • $50 - $150 • Remote

About The Position

Patronus AI is hiring a finance domain expert as an independent contractor to create realistic finance tasks that qualified professionals can solve with appropriate knowledge and tools, but that current AI models may get wrong. For each provided scenario, the contractor will create authentic working documents, write the task request, solve it, and define how the response should be graded. The resulting work will serve as evaluation and training data for frontier AI models. No AI background is required.

The work covers treasury operations, audit and assurance, trading and portfolio operations, or equity administration. Engagement is fully remote and asynchronous, with a commitment of at least 10 hours per week. Task authoring pays $50–$150 per hour, with payment when the complete task is submitted and a potential bonus for each accepted task. Strong performers may move up within the rate range or onto productivity-based pay. The trial task is paid, and work can begin on a rolling basis after passing it.

All tasks, documents, and solutions must be human-authored, original, and based on fictional scenarios; LLM-generated content and real client, employer, or personal data are not permitted. Quality is evaluated on accuracy, rubric validity, difficulty, and reviewer agreement.

Requirements

• At least 3 years of hands-on experience in one or more of treasury operations, audit and assurance, trading and portfolio operations, or equity administration.
• Working command of the relevant standards, controls, and rules in the area of expertise, such as US GAAP or IFRS, PCAOB/AICPA auditing standards and SOX, treasury policies and bank controls, trade compliance and restricted-list rules, or equity plan documents and fair-market-value requirements.
• Strong spreadsheet skills and the habit of tying every figure back to its source.
• Clear, precise writing in professional English.
• Ability to work independently and asynchronously against deadlines.
• Ability to produce original, human-authored tasks and solutions without LLM-generated content, using fictional scenarios rather than real client, employer, or personal data.

Nice To Haves

• CPA, CA, ACCA, CISA, CTP, CFA, or CEP credential.
• Experience at a Big Four firm, public company, bank, asset manager, or fund.
• Experience with relevant systems, such as SAP Treasury or another treasury management system, Bloomberg, AIM, EMSX, Carta, or Shareworks.
• Prior experience writing or reviewing AI evaluation tasks, exam questions, case studies, or training material.
• Familiarity with how AI models fail, including hallucinated figures, confident arithmetic errors, and misread documents.

Responsibilities

• Build realistic evidence files from provided scenarios, such as bank statements, cash positions, deal confirmations, settlement records, trial balances, GL extracts, audit workpapers, control evidence, trade blotters, rebalance instructions, position and cash reconciliations, cap tables, grant records, and board consents.
• Write clear task requests that specify the deliverable and explain how uncertainty and justified assumptions should be handled.
• Solve each task end to end and document a reference solution, including the reasoning behind each step.
• Write explicit scoring criteria that enable independent graders to reach the same verdict and include acceptable alternative answers.
• Run tasks against current models, identify where and why models fail, and adjust task difficulty.
• Review personal and peer tasks for ambiguity, errors, answers exposed in materials, and shortcuts that bypass intended reasoning.
• Incorporate reviewer feedback within agreed turnaround times and flag guideline gaps or edge cases.

What This Job Offers

• Career Level: Mid Level
• Education Level: No Education Listed