隐私政策
Privacy Policy
关于本政策
Envori 团队运营 Envori。本政策说明我们在提供消息沟通、事项协调及 AI Agent 辅助服务时如何处理信息。不同渠道和功能的可用情况可能不同；仅使用网站、通过某个渠道联系或参与某个事项，不代表同意把你的信息披露给其他参与人。
1. 我们处理的信息
- 消息与渠道身份：你发送的内容、姓名或显示名称、邮箱、电话号码或平台提供的账号标识，以及收发人、消息 ID、线程或回复引用、时间、渠道状态等元数据。
- 附件与原始渠道记录：消息可能包含图片、音频、文件、链接及附件元数据。渠道可能向我们提供附件地址或原始事件内容；我们可能保存这些原始记录以核对来源、处理不支持的内容和排查问题，并不意味着所有附件都会被下载、识别或交给 AI。
- 事项与授权：你的需求、参与对象、事项状态、你确认或撤回的可分享片段、接收人、用途、范围及相应确认记录。使用相关功能时，也可能处理公开资料、资料来源及观察时间、你提交的补充信息或团队提供的事项材料。
- Agent 与审核记录：提供给 AI 的相关输入和上下文、版本依据、系统可见的判断说明或输出、工具调用及结果、人工问题与答复、操作人及审核记录。这不表示我们取得或保存模型未向系统提供的内部推理。
- 发送与审计记录：待发送内容、发送请求、渠道受理或发送事实、可取得的回执、结果不明的尝试、错误、任务运行及安全审计记录。渠道受理、已发送、已送达、已读和对事项的同意是不同状态。
2. 使用目的与 AI、人工审核
我们使用这些信息来接收和回复消息、理解及澄清需求、协调事项、核对身份和授权、执行获准的工具动作、记录交接与发送结果，以及维护安全、可靠性和排查问题。获授权的团队还可能分析执行顺序、交接与故障，改进提示词、Skills（工作流程指引）及服务质量。相关信息可能提供给 AI 服务处理；获授权的 Envori 操作人员也可能为人工答复、核对不明引用或非文本消息、审核授权和处理发送异常而查看必要记录。AI 可能出错，我们可能要求进一步澄清或人工审核。
3. 向其他人披露
在服务中把你的信息交给其他参与人时，我们依据相应授权核对对象、用途和内容范围。普通来信、公开资料、资料修订、兴趣表达或渠道回执，本身不等于新的对外披露许可。你可以要求停止或撤回尚未执行的授权；已经送到其他人或渠道的信息不能保证被收回，对方可能保留自己的副本。
4. 服务提供商
我们使用服务提供商支持基础设施、AI 处理和通信。仅当你使用相应渠道或功能时，相关信息才可能涉及 Meta（Instagram）、AWS（托管与存储）、OpenAI Codex（AI Agent 处理）、Instantly（邮件渠道）、Google Workspace（邮箱）、Sendblue 或 Infobip（相应通信渠道）。渠道提供商也按其自身条款和隐私政策处理信息；本政策不替它们作出数据使用或模型训练方面的承诺。
当前服务使用 AWS 美国区域及获授权团队的运维环境；服务提供商可能在你所在地区之外处理信息。本隐私政策静态页不包含广告追踪脚本，也不设置 Cookie；访问基础设施可能为内容传输和安全处理 IP 地址、请求时间及浏览器信息。
5. 保存、安全与删除申请
我们按服务必要性及适用的授权、争议处理和法定要求评估保存范围和期限，目前不承诺统一固定保存期限。信息不再为合法业务所必要、收到有效删除请求或停止相关服务时，我们会及时评估并按适用要求处理删除；依法必须保留的部分另行说明。涉及 Meta 平台数据时，遵循适用平台更严格的删除要求。部分原始消息和审计记录采用追加留档，但这不表示可以无限期保留。我们采取身份验证、访问范围限制和发送前检查等措施，但不能保证任何系统绝对安全。
你可以申请访问、更正、停止使用或删除与你有关的信息。请求不需要 Envori 账号；即使没有平台登录，也可以通过邮件提出。请联系下方邮箱，说明相关渠道、账号和事项，避免发送不必要的敏感信息。我们会人工审核请求、核实身份及影响范围，并及时说明能够执行的处理、预计步骤及有具体依据需要保留的记录。删除不是即时或自动完成；相关副本、审计依据、第三方持有记录及技术依赖可能需要进一步处理，无法完成的部分会说明。停止某个事项也不等于删除全部历史记录。
6. 联系与更新
联系 Envori 团队：hello@byenvori.com。我们可能随服务变化更新本政策，并在本页标示更新日期；重大变化会通过适当方式通知。本页中英文描述同一处理做法；如有疑问，请联系我们澄清。
About this policy
Envori is operated by the Envori team. This policy explains how we handle information when providing messaging, matter coordination and AI Agent assistance. Availability varies by channel and feature. Visiting this site, contacting us through a channel or taking part in a matter does not by itself authorize disclosure of your information to other participants.
1. Information we process
- Messages and channel identities: content you send; names or display names; email addresses, phone numbers or platform account identifiers; and metadata such as senders, recipients, message IDs, threads, reply references, timestamps and channel states.
- Attachments and original channel records: messages may contain images, audio, files, links and attachment metadata. A channel may provide attachment URLs or original event content. We may retain these records to verify sources, handle unsupported content and investigate issues. This does not mean every attachment is downloaded, interpreted or provided to AI.
- Matters and permissions: your requests, participants, matter states, confirmed or withdrawn shareable excerpts, recipients, purposes, scope and confirmation records. Relevant features may also process public profiles, sources and observation dates, information you add, or matter materials supplied by a team.
- Agent and review records: relevant AI inputs and context, version references, explanations or outputs visible to our system, tool calls and results, questions and human answers, operator identities and review records. This does not mean we obtain or retain internal model reasoning that is not provided to the system.
- Sending and audit records: proposed messages, send requests, provider acceptance or sending facts, available receipts, uncertain attempts, errors, task runs and security audit records. Acceptance, sending, delivery, reading and consent to a matter are different states.
2. Purposes, AI and human review
We use this information to receive and reply to messages, understand and clarify requests, coordinate matters, verify identities and permissions, perform authorized tool actions, record handoffs and sending outcomes, and maintain security and reliability or investigate issues. Authorized teams may also analyze execution order, handoffs and failures to improve prompts, Skills (workflow guidance) and service quality. Relevant information may be processed by AI services. Authorized Envori operators may also review necessary records to answer questions, resolve uncertain references or non-text messages, review permissions and handle sending exceptions. AI can make mistakes; we may request clarification or human review.
3. Disclosure to other people
When sharing your information with another participant through the service, we check the recipient, purpose and content scope against the relevant permission. An ordinary message, public profile, profile change, expression of interest or channel receipt does not itself grant new permission to disclose information. You may ask us to stop or withdraw an authorization that has not yet been acted on. We cannot guarantee recall of information already sent to another person or channel, and recipients may retain their own copies.
4. Service providers
We use service providers for infrastructure, AI processing and communications. Information may involve Meta (Instagram), AWS (hosting and storage), OpenAI Codex (AI Agent processing), Instantly (email channel), Google Workspace (mailbox), Sendblue or Infobip (their respective communication channels) only when you use the relevant channel or feature. Channel providers also process information under their own terms and privacy policies. This policy does not make promises on their behalf about data use or model training.
The service currently uses an AWS US region and authorized team operations environments. Providers may process information outside your location. This static privacy page contains no advertising tracking scripts and does not set cookies. Access infrastructure may process IP addresses, request times and browser information for transmission and security.
5. Retention, security and deletion requests
We assess retention scope and duration based on service necessity and applicable permission, dispute and legal requirements. We do not currently promise one fixed retention period. When information is no longer necessary for a legitimate business purpose, a valid deletion request is received or the relevant service stops, we promptly assess and handle deletion as required, explaining any legally required retention. Applicable stricter Meta platform deletion requirements apply to Meta platform data. Some original message and audit records are append-only; this does not justify indefinite retention. We use measures including authentication, scoped access and checks before sending, but no system can be guaranteed completely secure.
You may request access, correction, cessation of use or deletion of information about you. No Envori account is required; you can submit a request by email even without a platform login. Contact the email below and identify the relevant channel, account and matter without including unnecessary sensitive information. We manually review requests, verify identity and assess the affected scope, then promptly explain the available handling, expected steps and specific grounds for any records that need to remain. Deletion is not instant or automatic. Copies, audit evidence, third-party records and technical dependencies may require further handling; we will explain any part that cannot be completed. Stopping a matter does not itself delete all historical records.
6. Contact and updates
Contact the Envori team at hello@byenvori.com. We may update this policy as the service changes and show the updated date here. Material changes will be communicated through an appropriate channel. The Chinese and English text describe the same practices; please contact us if anything needs clarification.