Persian AI Product Evaluation Starter Set
Version: 1.0
Published by: Sepehr Bayat
Canonical article: https://sepehrbayat.com/fa/blog/persian-ai-evaluation-set/

This file contains original, synthetic product-evaluation cases. It does not
copy ParsiNLU, PeKA, PK-BETS, ELAB, FarSense, PARSE, or any other academic
benchmark. It is a starter template, not a scientifically validated benchmark,
leaderboard, or representative sample of every Persian variety or product risk.

Before use:
- adapt cases to real product tasks;
- remove personal or confidential data;
- write acceptance criteria before viewing model outputs;
- use qualified native Persian reviewers;
- involve domain experts for medical, legal, financial, or other high-risk use;
- record model, prompt, tools, date, retries, and reviewer decisions;
- report results per category instead of hiding weak slices in one average.

License: Creative Commons Attribution 4.0 International (CC BY 4.0)
License URL: https://creativecommons.org/licenses/by/4.0/
Attribution: Sepehr Bayat, Persian AI Product Evaluation Starter Set, version 1.0.