Privacy notice for the hosted trial.

What the service receives

• An access-code session and basic session timestamps.
• Uploaded document names, hashes, file metadata, and document content needed to parse the comparison.
• Extracted candidates, evidence locations, review decisions, report snapshots, and purchase-intent contact details if you submit them.
• Anonymous product analytics such as page views, upload count, processing completion, evidence opens, exports, and purchase-intent events.

Why it is used

Documents and extracted records are used to run the comparison, show source evidence, preserve a report snapshot, and support deletion or troubleshooting. Analytics is used to understand whether people complete a comparison. Analytics events do not include document text, source quotes, comparison IDs, report IDs, or email addresses.

Model processing

When a model is configured for the service, selected text excerpts and document blocks may be sent to that configured OpenAI-compatible model provider so it can propose candidates. Deterministic rules check units, evidence, ownership, and conflicts afterward. When no model is configured, the demo extractor is used for flow testing.

Do not upload confidential, regulated, or personal data unless you have assessed the provider and deployment settings for your use case.

Retention and deletion

The hosted trial keeps documents and reports for 7 days by default, then cleanup removes them. The exact deletion time can vary if a job is still running or a provider is temporarily unavailable. To request early deletion or ask a privacy question, email llzclm@gmail.com with enough information to identify the session; do not send the document contents again.

中文说明

公开试用会处理访问会话、上传文件、解析结果、证据、报告和可选的购买意向联系方式。文件和报告默认保留 7 天，之后清理。配置模型时，部分原文摘录可能发送给你指定的 OpenAI-compatible 服务商；没有模型时使用演示抽取器。统计事件不包含原文、比较 ID、报告 ID 或邮箱。

如需提前删除或询问隐私事项，请联系 llzclm@gmail.com。在确认适用性前，请不要上传机密、受监管或个人敏感资料。