U.S based Accountant (CPA/CA)

What you would do

• Read tax statutes and regulations at the detailed level to identify all requirements an AI system must capture
• Create rubrics that list specific, testable criteria grouped by legal requirement with severity levels assigned
• Review AI translations of statutes and check them against your rubric criteria for pass or fail status
• Design test scenarios with realistic facts and edge cases that expose gaps in AI legal reasoning
• Provide structured written feedback on how AI outputs diverge from statutory requirements and source language

Who they want

• Active CPA license, Chartered Accountant designation, or proven track record as a U.S. tax attorney
• Demonstrated expertise in sales and use tax statutes or broader state and local tax regulatory experience
• Demonstrated skill parsing complex legal language and converting statutory requirements into verifiable specifications
• Comfort working with conditional logic, AND/OR structures, thresholds, exceptions, and defined legal terms
• Strong written communication focused on precision so another reviewer reaches identical conclusions from your criteria

Main skills

What the interview asks about

1. 1.Statute translation to testable criteria

   The core value is converting legal language into specifications; evaluators must confirm candidates understand statutory requirements deeply and can specify them unambiguously.

   For example: “You have read a state statute on nexus requirements for use tax collection that contains three conditional clauses and two defined exceptions. Walk through how you would structure your rubric and assign severity to each element.”

2. 2.Severity and materiality judgment

   Candidates must distinguish between critical and secondary requirements; treating everything as critical wastes AI training resources and obscures genuine gaps.

   For example: “A statute defines when a seller must collect use tax but includes a dollar-threshold exception and a separate carve-out for certain business categories. How would you weight these components in your rubric, and why?”

3. 3.Edge case and exception identification

   AI models often fail on exceptions and boundary conditions; evaluators need to know whether candidates think through what statutes exclude or when rules flip.

   For example: “You are building a test scenario for a statute that exempts certain resales but requires documentation. What specific facts would you include to verify the AI correctly handles both the exemption trigger and the documentation requirement?”

4. 4.AI translation evaluation rigor

   Evaluators must assess whether candidates can read an AI's legal output, compare it to source law, and identify what is missing or incorrect without being influenced by the AI's confidence.

   For example: “An AI translates a statute on credit application but omits a defined term from the source text. How would you determine if this is a critical gap or a simplification the AI can justify?”

5. 5.Written criteria clarity

   Other reviewers must be able to apply the rubric and reach identical conclusions; evaluators assess whether candidates write unambiguous criteria that don't rely on subjective judgment.

   For example: “Draft one rubric criterion from a statute that includes a threshold and a conditional qualifier. How would you phrase it so a junior accountant using your rubric reaches your same conclusion?”

A task you may get

Read a published state tax statute on a single topic, build a complete rubric with criteria and severity assignments, then evaluate a sample AI translation against your rubric and flag any mismatches.

How to prepare

• Study one recent state sales tax statute section and identify all its requirements, conditions, and exceptions by rereading the source text
• Review examples of legal writing that breaks down statutes into rules and practice converting those back into testable criteria
• Identify 2-3 ambiguous clauses from a tax statute and draft two different interpretations, then note which is most defensible given context

The facts

Pay

$65–70/hr

Open to

Bangladesh, Hong Kong, India, Indonesia, Japan, Kazakhstan, Kyrgyzstan, Malaysia, Pakistan, Philippines, Singapore, Sri Lanka, Taiwan, Thailand, Uzbekistan, Vietnam, Austria, Belarus, Belgium, Denmark, France, Germany, Greece, Italy, Netherlands, Portugal, Russia, Spain, Switzerland, United Kingdom, Argentina, Brazil, Chile, Colombia, Mexico, Peru, Algeria, Bahrain, Egypt, Iraq, Jordan, Kuwait, Lebanon, Libya, Morocco, Oman, Palestine, Qatar, Saudi Arabia, Tunisia, United Arab Emirates, United States, Canada, Nigeria, Kenya, South Africa, Ghana, Ethiopia

Field

Finance

Role type

Specialist

Posted

8/14/2026

Places left

10

We wrote this page from the public micro1 listing. It may be out of date, so read the full posting before you apply.