PJQv1.2

Privacy Policy

*Last updated: 2026-06-03*

PJQ (Public Judgment Quotient) is built on a "privacy-by-default" foundation. We do not build user dossiers, we do not store raw third-party comments, and we collect only the absolute minimum data required to run the platform.

1. Data We Collect from Users

To provide you with access to our services, we collect:

  • Email address — used exclusively for account creation and authentication.
  • API Key — automatically generated by our system.
  • Billing information — top-up (deposit) amounts in USDT equivalents and public transaction hashes (we operate solely on crypto-tips).
  • Session metadata — anonymized IP addresses and user-agents to protect our API from DDoS and abuse.

    • We DO NOT collect: your real name, phone number, physical address, or credit card information.

    2. Handling of YouTube Comments

    PJQ analyzes publicly available comments under videos. Our processing pipeline is strictly isolated:

    1. Comments are loaded into the server's operative memory (RAM).

    2. The platform aggregates the data (distribution across 7 categories, calculating mood and polarization indexes) and extracts 2-4 anonymized quotes as exemplars.

    3. Raw comment texts and their authors' profiles are permanently deleted. Only statistical aggregates and the selected exemplar quotes are saved to our database.

    3. Data Retention

  • Verdicts (statistics) and exemplar quotes: Stored indefinitely as part of PJQ's public knowledge base.
  • User Data (email, API key, API usage history): Retained as long as your account remains active.

    • 4. User Rights (Including GDPR)

    You have full control over your account. At any time, you can request the deletion of your account, email address, API keys, and analysis history. To do so, please send a request to `schowalterl235@gmail.com`. Data will be deleted within 30 days.

    5. Third Parties

  • YouTube Data API: We use this to read public metadata and video comments. We do not require linking your YouTube account and have no access to it.
  • Stance classification (our own model): The core classifier is a model hosted on our own servers — individual comments are not sent to any external AI vendor for classification.
  • Third-party AI API (DeepSeek): For auxiliary production tasks — detecting the topic/domain and generating the video summary card — we send public video metadata (title, tags, public context) to the DeepSeek LLM API. If you explicitly enable the optional QA re-check, short snippets of public comment text may also be sent there for a second opinion (this is off by default). We never send your account data (email, API key, wallet) to any AI vendor.
  • Offline AI Providers (Anthropic): Used only in a secure offline environment to generate training datasets for our local model — never in the live request path.

    • 6. Cookies

    We use only technical cookies (necessary for maintaining your authentication session and UI preferences). We do not use advertising trackers or third-party analytics. For details, see our Cookie Policy.

    7. Contact Us

    If you have privacy-related questions, please contact us at: `schowalterl235@gmail.com`.