What the work involves
Day to day, you are producing and judging accounting content that an AI model will be measured against. Typical tasks include drafting realistic prompts a client or controller might ask, writing the reference answer with the authoritative citation (ASC topic, IRC section, IRS instruction), and grading two or more model responses against a rubric — flagging where the model invented a rule, applied a superseded standard, botched a journal entry, or reached the right number with wrong reasoning. Some projects lean toward tax memos and return logic; others toward US GAAP treatment, reconciliations, audit workpaper review, or building small fact patterns with a defensible single correct answer.
The hardest part is not the accounting — it is writing down why an answer is wrong at a level of specificity a reviewer can audit. "Incorrect revenue recognition" gets sent back; "applies ASC 606 step 4 allocation to a contract that fails step 1 because collectibility is not probable" does not.
What the platform screens for
micro1 runs an AI-led interview before any project match. Expect it to probe:
- An active U.S. CPA licence or equivalent CA credential, and which state board issued it
- Where your depth actually sits — federal tax vs. US GAAP technical accounting vs. audit vs. controllership
- Whether you can cite standards from memory and stay consistent when the interviewer pushes back on a correct answer
- Written English precise enough for rubric authoring, not just client email
- Honest weekly availability, since most projects run in short intense batches
Logistics
Fully remote and largely asynchronous, with occasional calibration calls on U.S. business hours. Volume is not guaranteed and fluctuates by project; most contributors treat this as 5–20 hours a week alongside practice or a full-time role. Pay is hourly on tracked task time at the band observed for this listing, not a salary, and rates can differ by project difficulty.