FICTIONAL RECORDS · ALL CASES NOT RUN
Matching figures: what did the review establish?
A companion to TaxFlo’s November 3 field note. This twenty-minute exercise separates receipt, consistency and issuer-origin questions. It is not an authenticity or fraud-detection certification.
Before starting
Use only these invented records in an approved demonstration environment. Agree on the reviewer’s permission and what the status means. Start with the firm’s existing process. Keep the facilitator key separate until the reviewer records an answer.
Twenty-minute sequence
- 0–5 minutes: receive organizer-1 and statement-v1. Identify the source/version and expected comparison.
- 5–10 minutes: compare the identical 12,500 amount. Record what was checked and what remains unknown.
- 10–15 minutes: introduce statement-v2 and explanation-1. The account reference changes; the amount does not. Ask what requires reconsideration.
- 15–20 minutes: give the handoff to a second authorized reviewer without an oral explanation. Ask what the status permits them to conclude, what remains open and who owns it.
Observation sheet
Every row begins NOT RUN. Fill actual observations only after authorized execution. Count unsupported conclusions, missing explanations and verification effort; preserve appropriate unknowns.
| Case | Test | Status | Observed result / owner |
|---|---|---|---|
| R1 | Receipt only | NOT RUN | ________________ |
| C1 | Same amount | NOT RUN | ________________ |
| V1 | Correction, same amount | NOT RUN | ________________ |
| H1 | Blind authorized handoff | NOT RUN | ________________ |
Files
- Fictional input records (JSON)
- Editable observations (CSV)
- Facilitator key — separate from first review
Professional boundary
Tax judgment, source-verification decisions and final approval remain with the practitioner. This packet announces no TaxFlo integration, product test result, savings or customer outcome. Raman Yakkala is AmitaSoft’s founder; TaxFlo is built on SafeVault’s governed document foundation.
Background: exploratory loan-document preprint and limitations. Three reviewers and no genuine-file controls do not establish CPA detection accuracy. The exercise and inputs above are original, proposed editorial material.