English
1. Who we are
SENSE is operated by 北京衍数科技有限公司 (Beijing Yanshu Technology Co., Ltd., "Yanshu Tech"). Contact: service@yanshutech.com, 创业中路32号楼, 海淀区, 北京市, China. The parent website is yanshutech.com.
2. Data we process
- Camera video frames and microphone audio — captured in your browser only while a session is running. They are streamed over an encrypted connection (HTTPS/TLS) to the analysis servers to extract behavioral, acoustic and physiological features in real time.
- Biometric features — derived measurements such as facial action-unit scores, gaze and blink estimates, voice pitch, jitter, shimmer, HNR and estimated heart rate. Under laws such as China's PIPL, these may qualify as sensitive personal information.
- Session content — the background information you provide, the questions generated, transcribed answers, and the resulting report.
- Account data — username, email address, login credentials and credit/payment records (payments are processed by WeChat Pay).
- Report delivery data — the recipient email address you enter when creating a remote link.
- Technical data — IP address, browser type, timestamps and error logs, used for security and reliability.
3. How video and audio are handled
- Video and audio are processed in real time to extract the features listed above.
- The application backend does not write raw video or audio to persistent storage; session state is held in memory and released when the session ends.
- Transport is encrypted with standard TLS. No internet transmission is 100% secure.
- Do not use SENSE to record sensitive, classified or third-party personal content, and do not upload documents containing personal data into the context engine.
4. Third parties
- Large language model — the background information and transcribed answers are sent to our own self-hosted LLM inference service (qwen3-32b) to generate questions and the final report summary. They are not sent to a third-party LLM provider.
- Payment provider — WeChat Pay processes top-up payments; we do not store card or bank credentials.
- Email delivery — report and verification emails are delivered through Tencent Exmail (腾讯企业邮).
- Speech recognition — the SenseVoice speech model runs on our own analysis servers; audio is not sent to a third-party speech API.
We do not sell personal information.
5. Retention
- Session state and session history — released when the session ends and no longer retained after you log out; raw video and audio are not written to persistent storage.
- Account, credit and billing records — retained after registration and not deleted automatically. You may request account deletion by email; we will delete the records after verifying your identity.
- Emails — reports and verification codes remain in the recipient's mailbox under the mail provider's terms.
6. Your rights
You may request access to, correction of, or deletion of your personal information by emailing service@yanshutech.com. Where biometric data is processed, we rely on your explicit consent, which you give by choosing to start a session; you can withdraw it at any time by stopping use and requesting deletion. You may also lodge a complaint with your local data-protection authority.
7. Children
SENSE is not intended for children under 14. Do not start a session involving a child without verifiable guardian consent.
8. Changes
We may update this policy as the service evolves. The "last updated" date at the top of this page reflects the current version.
中文
1. 我们是谁
SENSE 由北京衍数科技有限公司运营。联系方式： service@yanshutech.com， 北京市海淀区创业中路32号楼。公司官网：yanshutech.com。
2. 我们处理哪些数据
- 摄像头视频帧与麦克风音频 —— 仅在会话进行期间于浏览器中采集，并通过加密连接（HTTPS/TLS）传输至分析服务器，用于实时提取行为、声学与生理特征。
- 生物特征数据 —— 由上述信号派生的测量值，包括面部动作单元分数、视线与眨眼估计、声音基频、jitter、shimmer、HNR 以及估计心率。根据《个人信息保护法》等法律，此类信息可能属于敏感个人信息。
- 会话内容 —— 你填写的背景信息、生成的问题、转写的回答与最终报告。
- 账号数据 —— 用户名、邮箱、登录凭证与积分/支付记录（支付由微信支付处理）。
- 报告接收信息 —— 创建远程链接时填写的接收邮箱。
- 技术数据 —— IP 地址、浏览器类型、时间戳与错误日志，用于安全与稳定性保障。
3. 视频与音频如何处理
- 视频与音频实时处理，用于提取上述特征。
- 应用后端不会将原始视频或音频写入持久化存储；会话状态保存在内存中，并在会话结束时释放。
- 传输采用标准 TLS 加密。没有任何互联网传输是 100% 安全的。
- 请勿使用 SENSE 录制敏感、涉密或他人隐私内容，也请勿在上下文分析引擎中上传含个人数据的文件。
4. 第三方
- 大语言模型 —— 背景信息与转写回答会发送至我们自建的 LLM 推理服务（qwen3-32b），用于生成问题与最终报告摘要；不会发送给第三方 LLM 服务商。
- 支付服务商 —— 充值由微信支付处理；我们不存储银行卡或支付凭证。
- 邮件服务 —— 报告与验证码邮件通过腾讯企业邮发送。
- 语音识别 —— SenseVoice 语音模型运行在我们自己的分析服务器上，音频不会发送至第三方语音 API。
我们不会出售个人信息。
5. 数据保留
- 会话状态与会话历史 —— 在会话结束时释放，退出登录后不再保留；原始视频与音频不写入持久化存储。
- 账号、积分与计费记录 —— 注册后持续保留，不会自动删除。你可以通过邮件申请注销账号，我们将在核实身份后删除相关记录。
- 邮件 —— 报告与验证码将保留在收件人邮箱中，适用邮件服务商的条款。
6. 你的权利
你可以通过 service@yanshutech.com 请求查阅、更正或删除你的个人信息。 涉及生物识别信息处理时，我们依赖你的单独同意——你通过主动开始会话给予同意，并可随时停止使用并请求删除以撤回同意。 你也可以向当地数据保护监管机构投诉。
7. 未成年人
SENSE 不面向 14 周岁以下儿童。未经监护人可验证的同意，请勿让儿童参与会话。
8. 政策更新
随着服务发展，我们可能更新本政策。页面顶部的“最后更新”日期对应当前版本。