Privacy Policy

Last updated September 24, 2026. This is an interim policy and will be replaced by a fuller one before HeyJanet begins charging.

Where your documents are

In a Google Shared Drive that we operate, in a folder belonging to your company. Your documents are held by us, on Google's infrastructure, under our Google Workspace account — not in a Drive of your own. We want that stated plainly rather than implied away, because it is the honest answer to “who is holding our files?”

What that means in practice:

• There is one copy, and it is the one in that folder. We do not duplicate your files into a database of ours. Our database holds only the metadata that makes a library trustworthy: titles, document types, version numbers, who filed what and when, and links back to the files themselves.
• The Google account you nominate is granted read access to your folder, which is how your AI assistant opens a document as it answers. Nobody outside that grant can reach it by link.
• You can download or copy anything, at any time, through Google Drive directly.
• You can ask us to hand the files over — see Leaving.

What we hold about people

• The name and email address of each person with access to a workspace, held on our behalf by our authentication provider.
• An activity record of what was uploaded, replaced or published, and by whom. This is the compliance asset — it is the thing that answers "which version was in force in March" — and it is kept deliberately.
• Company details you enter yourself: address, contact numbers, licence and insurance numbers, and the people responsible for each department.

Where content goes when you upload it

This is the part worth reading closely.

The text of a document you upload is sent to our AI classification provider (Anthropic) to work out where it belongs — which department, which document type, what to call it, and a one-line description of what it is for. That text is processed to produce those answers and is not retained by them for training.

Uploads are also screened before anything is stored, to refuse documents containing health information, Social Security numbers or payment card numbers. That screening happens on our servers.

Separately, your own AI assistant reads your library. That runs in your account with Anthropic, OpenAI or Perplexity, under your agreement with them, using access you grant and can revoke. We do not see those conversations.

Who else we use

• Google — Drive and Sheets, where your documents and index live.
• Clerk — sign-in and accounts.
• Neon — the database holding metadata, not files.
• Vercel and Netlify — hosting.
• Anthropic — document classification, as above.

What we do not do

We do not sell your data. We do not use your documents to train AI models. We do not share your content with anyone except the providers listed above, and only to run the service.

Leaving

Your documents leave with you. Because we hold them, getting them back is a step we have to take rather than something that happens by itself — so here is exactly what we do. On request we will either transfer the files into a Google Drive you control, or export the whole library to you, including the index and the version history. Your choice, at no charge, and we will not hold it up.

We then delete the metadata we hold, keeping only the minimum needed for our own records and legal obligations.

Contact

Questions, or a request about your own data: amir@fivespotmarketing.com.