What the work involves
You build the test, the answer, and the grading sheet. A typical task starts with you constructing a realistic business scenario — a portfolio company margin decline, a market entry question, a capital allocation decision — and assembling the multi-file input pack a real analyst would receive: messy spreadsheets, a PDF annual report, raw source data. You then produce the gold-standard deliverable yourself, at the quality level a partner or investment committee would actually accept: a structured PowerPoint deck with clean visual storytelling, a memo with a defensible recommendation up front, or an Excel model built from scratch with working scenario and sensitivity logic.
The second half of each task is evaluative. You write a detailed rubric — often 35+ discrete line items — covering MECE structure, arithmetic and sourcing accuracy, logical consistency between exhibits and conclusions, and executive-readiness. You then document how AI outputs fail: recommendations that hedge, issue trees that overlap, numbers that appear in a chart but nowhere in the supporting tab. That failure-mode documentation is often the most valuable artifact you produce.
What the platform screens for
micro1's screening is AI-led and specific. Expect to be asked which decks you personally owned versus contributed to, who reviewed them, and what changed after review. Expect follow-ups on modeling mechanics — how you structured a three-statement build, how you handled circularity, what your sensitivity toggles actually drove. Vague consulting vocabulary is a poor signal here; concrete recall of a single engagement is a strong one. Writing quality is assessed directly, since nearly every deliverable is executive-facing prose or a slide headline.
Logistics
- Contractor engagement, fully remote, NAMER and Europe time zones
- Fully async; no standing meetings, though calibration sessions may be offered
- Output-based pay — you're paid per accepted task, so effective hourly rate depends heavily on your speed with Excel and PowerPoint
- A weekly minimum task submission applies; treat this as a real commitment, not overflow work
- Roles typically fill within 48 hours, with first tasks expected 24–48 hours after onboarding
- No prior AI or ML experience required or expected