Privacy policy
Effective 2 November 2026.
Standup Bot is a private team tool operated by Marcus Tan. This policy explains how the bot handles meeting information.
What we access
With your authorization, the app can view events on all your calendars and see all your Google Docs documents. These are the permissions available through the Google APIs it uses.
The configured workflow searches the primary calendar for the standup, reads its event details and notes attachment, and reads the document's tabs. The operator can also provide a document ID directly.
Notes may contain transcripts, names, attendance, decisions, commitments, and other information discussed in the meeting. The bot also reads existing Notion tasks to match new work with earlier work.
How the information is used
The bot uses meeting information to extract summaries, decisions, action items, owners, due dates, and progress updates. It uses existing tasks to avoid duplicate work and identify items needing attention.
The resulting plan updates Notion and produces a Telegram recap. The Google permissions permit reading the source material; the task and meeting changes occur in Notion.
Which services receive it
- Google provides the calendar events and original notes documents through the accounts and permissions you authorize.
- Anthropic or DeepSeek receives meeting text, the people roster, and existing task context to extract the meeting plan. Claude is the primary model. DeepSeek is the configured fallback if Claude fails.
- Notion receives generated meeting summaries, decisions, attendee names, task details, updates, and links to the source notes.
- Telegram receives the recap, task counts, attention items, and Notion link in the selected chat. Private failure alerts can include a meeting title and source link.
- Cloudflare Pages hosts these information pages and may process visitor connection metadata to deliver and protect them. The bot does not send meeting notes or credentials to this website host.
Access to generated Notion pages and Telegram messages follows the permissions of the selected workspace and chat. The operator must select recipients who may receive that meeting information.
Storage and retention
The operator's Mac stores Google sign-in credentials, service credentials, and a delivery journal. The journal contains recap text, counts, document identifiers, meeting metadata, and delivery receipts. Private configuration and state files use restricted file permissions.
The bot code does not create a separate archive of raw transcripts. The original notes remain in Google. Generated content remains in Notion, Telegram, and any model-provider systems according to those services' settings and terms.
The current bot has no automatic deletion schedule for the local delivery journal. The operator retains it while needed for delivery and duplicate prevention, then removes it manually. Deleting an original Google document does not delete generated copies in other services.
Limits on use
The bot uses Google data to provide the standup workflow described here. The operator does not sell Google user data or use it for advertising. The bot does not train its own models.
The operator must follow the Google API Services User Data Policy, including applicable Limited Use requirements, for Google data and information derived from it.
Your choices
You can stop future Google access by removing Standup Bot from your Google Account connections. The operator can also stop the local scheduled job.
To remove local bot data or generated meeting content, contact Marcus. Existing Notion pages and Telegram messages need separate deletion in those services. Revoking Google access does not remove content already sent to them.
Before authorizing a meeting, ensure you have permission to process its content and share the results with the configured services and recipients.
Questions and changes
Contact marcus.tanchiyau@gmail.com about access, deletion, or this policy.
If the bot's data practices change, the operator will update this policy and obtain any additional authorization required before using data for a new purpose.