Domain Expert, AI Evaluation Task Writer: Finance (Contract, Remote)
关于 Patrus AI,PatronusAI 是一个开发模拟研究和基础设施的边界实验室,以加速人类对 AGI 的进展。我们在任务上模拟所有世界智能。我们是金融银行,林克斯,简单安全测试,版权捕捉器,人文最后考试等 AI 评估领域最早和影响力最大的研究团队。我们以前是 Meta AI,亚马逊 AGI 和谷歌等公司的 AI 研究人员和工程师。我们的客户包括基金会模型实验室和 Adobe 等财富500企业。我们得到了 Lightspeed 风险投资者,著名资本,斯坦福大学,诺姆布朗,戈古尔·拉贾兰等顶级投资者的支持。一看 · 承诺:独立承包商,完全远程,异步 · 付费:50-150美元/小时创作任务,支付当完成完整的任务提交,加上每项被接受的任务的潜在奖金。强大的业绩人员在该范围内或基于生产力的薪水上升。试验任务是支付的。· 承诺:10+个小时 /周 · 开始:滚动,你通过试验任务 · 地点:远程角色 你将设计一个合格的专业人员可以用正确的知识和工具来解决现实财务任务,而当今的人工智能模型会错误。每个任务都围绕着我们提供的场景构建。你创建的工作文件,你的角色中的某个人实际上处理,写出请求,自己解决并定义答案如何得到评级。你的工作成为边界模型的评估和培训数据。你不需要人工智能的背景。你必须是抓住数量但不应该的金融专业人士。我们聘请领域:现金定位和流动性管理,日常现金运营和银行调整,收益,风险和不一致的财政分析以及与对方确认和结算记录相匹配的预订交易 ·审计和保证:外部,内部或政府审计,技术保障,SOX 和 ITGC 控制测试,从证据和通过工作文件进行通行,例外,发现和补救 ·交易和投资组合操作:交易前遵守和清理,执行交易从重平衡说明,票据和片,并将位置,现金和持有权与指示和约束相比调整 ·股权管理:股票计划管理 (补贴,资产权授予,批准)
查看英文原文
About Patronus AI
Patronus AI is a frontier lab developing simulation research and infrastructure to accelerate progress toward human-aligned AGI. We are on a mission to simulate all of the world’s intelligence.
We are the team behind some of the earliest and most influential research in AI evaluation likeFinanceBench, Lynx,SimpleSafetyTests,CopyrightCatcher,Humanity’s Last Exam, and more. We are formerly AI researchers and engineers from companies like Meta AI, Amazon AGI, and Google. Our customers include foundation model labs and Fortune 500 enterprises like Adobe. We are backed by top-tier investors like Lightspeed Venture Partners, Notable Capital, Stanford University, Noam Brown, Gokul Rajaram, and more.
At a glance
- Engagement: Independent contractor, fully remote, asynchronous
- Pay: $50-150/hr for task authoring, paid when the full task is submitted, plus a potential bonus for each accepted task. Strong performers move up within that range or onto productivity-based pay. The trial task is paid.
- Commitment: 10+ hours per week
- Start: Rolling, as soon as you pass the trial task
- Location: Remote
The role
You'll design realistic finance tasks that a qualified professional could solve with the right knowledge and tools, and that today's AI models get wrong. Each task is built around a scenario we provide. You create the working documents someone in your role would actually handle, write the request, solve it yourself and define how the answer gets graded. Your work becomes evaluation and training data for frontier models.
You don't need an AI background. You do need to be the kind of finance professional who catches the number that ties out but shouldn't.
Areas we're hiring for
- Treasury operations: cash positioning and liquidity management, daily cash operations and bank reconciliation, treasury analysis of accruals, exposures and discrepancies, and matching booked deals to counterparty confirmations and settlement records
- Audit and assurance: external, internal or government audit, technology assurance, and SOX and ITGC controls testing, from evidence and walkthroughs through workpapers, exceptions, findings and remediation
- Trading and portfolio operations: pre-trade compliance and clearance, execution trading from rebalance instructions, tickets and blotters, and reconciling positions, cash and holdings against instructions and constraints
- Equity administration: stock plan administration (grants, vesting, approvals and plan availability), equity plan analysis (share reserves, outstanding awards and fair-market-value inputs), and corporate paralegal work on board consents and equity transaction documents
What you'll do
- Build realistic evidence files from a provided scenario: bank statements and cash positions, deal confirmations and settlement records, trial balances and GL extracts, audit workpapers and control evidence, trade blotters and rebalance instructions, position and cash reconciliations, cap tables, grant records and board consents
- Write clear task requests that state what the deliverable is and how uncertainty and justified assumptions should be handled
- Solve each task end to end and document a reference solution with the reasoning behind every step
- Write explicit scoring criteria that let two independent graders reach the same verdict, including acceptable alternative answers
- Run tasks against current models, find where and why they fail, and adjust difficulty
- Check your own and peers' tasks for ambiguity, errors, answers exposed in the materials and shortcuts that skip the intended reasoning
- Incorporate reviewer feedback within agreed turnaround times and flag guideline gaps or edge cases you run into
Who we're looking for
Required
- 3+ years of hands-on experience in one or more of the four areas above: treasury operations, audit and assurance, trading and portfolio operations, or equity administration
- Working command of the standards, controls and rules in your area (for example US GAAP or IFRS, PCAOB/AICPA auditing standards and SOX, treasury policies and bank controls, trade compliance and restricted-list rules, 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
Nice to have
- CPA, CA, ACCA, CISA, CTP, CFA or CEP
- Big Four, public company, bank, asset manager or fund experience
- Experience with the systems used in your area, such as SAP Treasury or another treasury management system, Bloomberg, AIM or EMSX, or Carta or Shareworks
- Prior experience writing or reviewing AI evaluation tasks, exam questions, case studies or training material
- Familiarity with how AI models fail (hallucinated figures, confident arithmetic errors, misread documents)
Our standards
- All tasks, documents and solutions are human-authored. LLM-generated content is not permitted.
- Tasks and solutions must be original. Nothing may be copied from published or easily searchable sources.
- Never use real client, employer or personal data. All scenarios are fictional.
- Quality is judged on accuracy, rubric validity, difficulty and reviewer agreement, not volume.
How hiring works
- Short AI interview and resume review
- One paid trial task, scored against a fixed quality checklist
- Onboarding and your first project
We aim to make a decision within two weeks of your application.
Patronus AI is an equal opportunity employer. We celebrate diversity in our workplace, and all qualified applicants will receive consideration for employment without regard to age, ancestry, color, family or medical care leave, gender identity or expression, genetic information, marital status, medical condition, national origin, physical or mental disability, political affiliation, protected veteran status, race, religion, sex (including pregnancy), sexual orientation, or other legally protected characteristics.
By clicking ‘Apply’, you agree to Greenhouse's Terms of Service and Privacy Policy.
By clicking 'Apply', you agree to Patronus AI, Inc. Privacy Policy.
Originally posted on Himalayas