Privacy
Privacy is documented around the data WatchLLM actually stores.
WatchLLM stores project, brand, prompt, response, citation, crawler, visitor, report, and team data needed to run the product loop.
Project data
Crawler logs
Visitor events
Local safe mode
Data used by the product
Workspace data can include organization membership, projects, brand books, competitors, personas, prompts, monitor runs, model responses, citations, generated content, actions, reports, crawler logs, and visitor pageviews.
Local default behavior
The development build uses deterministic local providers by default. It does not send prompts, content, report emails, team invite emails, object-storage uploads, or Google OAuth requests unless the corresponding live provider gate is explicitly configured.
Visitor and crawler analytics
Visitor tracking stores pageview analytics from the project tracking key, including normalized referrers and device/browser metadata. Crawler ingestion stores request metadata and AI crawler classification for project-owned logs.
Deletion and exports
The current product surfaces scoped deletes and CSV/report downloads where implemented. Production data-retention and customer-request workflows should be finalized before a hosted launch.
Review the shipped product surface.
The public docs and workspace show the local controls that back these trust statements.