Privacy and your data
What we store, why we store it, and how to get it back or get rid of it.
What we store
The pages we crawl from the site you point us at, the chunks and embeddings derived from them, the questions your visitors ask and the answers your bot gives, any answers you write yourself, and the email addresses visitors leave when they ask to be contacted.
Why
Each of those exists to answer a question the product would otherwise have to guess at: what your business does, what has already been asked, what you have already corrected, and who is waiting on a reply from you.
Who else sees it
Your content and your visitors' questions are sent to Google's Gemini API to generate embeddings and answers. Emails are sent through Resend and text messages through Twilio. Payment details are handled by Lemon Squeezy, our merchant of record, and never touch our servers. We do not sell data, and we do not use one customer's content to improve another customer's bot.
Exporting everything
A single request returns every row we hold for your organization as JSON in a zip: your bots, sources, crawled documents, extracted facts, saved answers, conversations, individual messages, gaps and outcome events. The numeric embedding vectors are excluded — they are a derived artefact of the embedding model and useless outside it — but the text they were generated from is included.
Deleting everything
Deletion takes effect immediately: the moment you confirm, your organization is unreadable everywhere in the product. The underlying rows are held for seven days so that an accidental deletion can be undone, then permanently destroyed by an automated job. After that we cannot recover them, because they no longer exist.
Visitors on your site
Visitors are identified by a random id stored in their own browser, not by anything personal, unless they choose to leave an email address. That id lets a conversation continue across page loads and nothing else.