What the work involves

You will build and review the evaluation material that AI labs use to measure financial reasoning. In practice this means authoring realistic prompts — a three-statement model that doesn't tie, a DCF with a mismatched discount rate and cash flow definition, an LBO with a debt schedule that breaks a covenant, a purchase-price allocation question — then documenting the correct answer with the reasoning a reviewer can verify. A second stream of work is grading: you read model outputs side by side, decide which is stronger, and write a short rationale that points to the specific error rather than a general impression. Some assignments ask you to red-team a model by finding the questions where it produces confident, well-formatted nonsense.

What the platform screens for

micro1 runs an AI-led interview before any human reviews your file. Expect it to probe the specifics of your experience: which sectors you covered, which models you built from scratch versus inherited, how you handled a particular adjustment. Vague seniority claims tend to collapse under follow-up questions, while precise answers — even about small deals — do well. The screen also tests evaluation judgment: whether you can separate a wrong answer from a right answer reached by weak logic, and whether you can articulate a grading standard another expert could apply consistently.

Typical backgrounds that clear the bar:

  • Investment banking, private equity, or corporate development experience with live models
  • FP&A, treasury, or controllership roles with technical accounting exposure
  • CFA charterholders, CPAs, or MBA finance concentrations with practitioner work behind them
  • Valuation, transaction advisory, or equity research analysts

Logistics

Fully remote and largely asynchronous. Most contributors work 10–20 hours a week around a day job, though some projects offer more volume for short bursts. Assignments arrive in batches with deadlines measured in days rather than hours, and pay is hourly within the observed $50–100 band — actual offers vary by credential depth, project, and whether you are authoring or reviewing. Volume is project-driven and not guaranteed week to week.