Privacy Policy

Baby LodeStar

Effective date: August 2, 2026 Last updated: August 2, 2026

Baby LodeStar is built by lolpick ("we", "us"). This policy explains what the Baby LodeStar iOS app does with your information.

The short version. Baby LodeStar stores your baby's sleep records on our servers so the app works across launches and reinstalls, and so an AI model can read them to draft a daily plan. It creates an account for you automatically, without asking for an email address or a password. It contains no analytics, no advertising and no tracking of any kind, and you can erase everything — including the account — from inside the app.

1. The account we create for you

The first time you open Baby LodeStar, the app signs in anonymously. This creates an account identified only by a random identifier. We never ask for, and never receive, your name, email address, phone number or password.

That account is what separates your baby's records from everyone else's. It is stored on your device in the iOS keychain. If you delete the app, that key is what is lost — which is why the app also offers an explicit deletion option (section 7).

2. What you enter, and where it goes

To produce a schedule, Baby LodeStar holds:

• your baby's first name, date of birth, and the situation you chose at sign-up ("nights are broken", "naps are a fight", "just want to keep track")
• sleep records — when each sleep started and ended, whether it was a nap or night sleep, and any note you add
• the daily courses the app drafts, and any changes you make to them
• your conversation with the coach
• app settings — briefing time and the reminder toggles

This is stored on servers we operate through Supabase, hosted in the United States (Ohio). Each account can only read and write its own rows; this is enforced by the database itself, not only by the app.

3. The AI planning feature, and the one third party that sees your data

When the app drafts a daily course or you send a message to the coach, we send the information needed to answer to OpenAI, which runs the language model.

What OpenAI receives: