Financial reasoning built by working analysts.
Finance models fail on the fourth step of a calculation, not the first. We commission vetted analysts and finance professionals to build and grade the data they learn from.
Corporate finance, with proof.
Models can define enterprise value and still get a valuation wrong. Common failures include mixing fiscal and calendar periods, treating a non-GAAP figure as reported, double-counting cash in a bridge, flipping a sign in a cash flow statement, or carrying a small arithmetic error through a multi-step model. HUMXN commissions credential-checked analysts, bankers and FP&A professionals to build models, write problems with worked solutions and grade model reasoning step by step.
Each item is created to your specification, fingerprinted at intake, checked for duplicates and reviewed by a second expert. Spreadsheets and documents are scanned for personal information and masked, and label provenance records whether a judgment came from the author, a model suggestion or an expert reviewer. Deliveries come with a signed receipt and audit trail, in JSONL, CSV, Parquet or Croissant, so the data drops into your training or evaluation pipeline.
For finance professionals, this is paid, remote work that uses the modeling and analysis you already do. You build models, write problems, grade model answers and review a colleague's work, at a rate shown before you accept each job. The work creates data for AI systems and is not investment advice to anyone. Your identity is not published, and approved work is paid automatically through Stripe after identity and tax verification.
Built to your specification.
Every engagement starts from a written spec and a pilot batch. These are the most common requests we source for corporate finance.
Valuation models
DCF, comparable company and precedent transaction models built from scratch, with assumptions and every formula documented.
Problem sets with solutions
Graduate and professional-level problems in valuation, capital structure, M&A and LBO math, each with a worked solution.
Filings analysis
Expert-labeled extractions and analyses of annual reports and financial statements, with periods and adjustments made explicit.
Step-level reasoning grades
Analysts grade each step of a model's calculation, marking where an error entered and how it carried through.
Preference judgments
Comparisons of model answers on accuracy, clarity of assumptions and appropriate caveats, with written rationales.
Spreadsheet agent tasks
Tasks and reference solutions for agents that build or audit financial spreadsheets, checked cell by cell.
01Errors compound
A small mistake early in a model distorts every output. Step-level grading shows exactly where reasoning went wrong.
02Periods and definitions
Fiscal years, trailing periods and non-GAAP measures are easy to mix up. Expert-built items make each choice explicit.
03Contaminated benchmarks
Textbook problems are widely published. Newly written, fingerprinted problems give evaluation results you can rely on.
Paid work in your field.
Remote and flexible, with the rate stated before you accept. Every job is reviewed by a second expert, and approved work is paid automatically.
- A degree in finance, economics, accounting or a quantitative field
- Several years in banking, research, private equity, corporate development or FP&A
- CFA charter or progress through the program, or an MBA with a finance focus
- Hands-on experience building three-statement, DCF or LBO models
- Sector specialisms such as energy, banks, insurance or technology
Build reference models
Construct valuation or operating models from a brief, documenting assumptions and formulas.
Write problems with worked solutions
Draft professional-level finance problems and step-by-step answers that a model should reproduce.
Grade model reasoning
Mark each step of a model's calculation and identify where and why it went wrong.
Review another expert's work
Check a colleague's models and grades as second reviewer before delivery.
Corporate finance: asked often.
What corporate finance data can you deliver?
Reference valuation and operating models, problem sets with worked solutions, filings extractions and analyses, step-level grades of model reasoning, preference judgments and tasks for spreadsheet agents. Each request is scoped by topic, sector and difficulty, then matched to finance professionals verified in that area. A pilot batch confirms the standard before scaling.
Who builds the models?
Credential-checked finance professionals: banking and research analysts, private equity associates, corporate development and FP&A staff. Each passes identity and credential checks, and may complete a short, paid assessment. Every item they deliver is reviewed by a second expert, and label provenance records who created and who confirmed it.
Is this financial advice?
No. Experts create and review data for AI systems. Nothing they write is a recommendation to any investor, and no advisory relationship is formed. Problems and models may use real public companies or hypothetical ones, depending on your specification, and are used only as training or evaluation data.
How do I get paid?
Rates are set by field and task and shown before you accept a job. Once a second expert approves your work, it is credited to your balance and paid automatically through Stripe after your identity and tax details are verified. Every job appears on an itemized ledger.
How much time is required, and can I work around my job?
There is no minimum commitment. Opportunities are offered when they match your verified field, each with a stated scope and deadline, and you choose which to accept. Work is remote, so you can fit it around your current role. Check your employer's policies on outside work before accepting.
Can we get exclusive rights?
Yes. Commissioned finance datasets can be licensed exclusively, with scope, term and exclusivity agreed in writing before work begins. Standing orders can keep new problems and models arriving at the same standard, which is useful for refreshing held-out evaluation sets.
Corporate finance, done by experts.
Need expert-made data in this field, or have the expertise to make it? Start here.