Privacy Policy

This is first because it is the reason people hesitate to type anything true into an app like this.

• Your employer is never told you use this, and never shown anything you write here. Nobody has an employer view.
• Nothing you write is sold or used for advertising.
• Nothing you write is used to train an AI model. The text goes to OpenAI through the interface they keep out of training.
• Nobody else using the app can see your practices. The database enforces this, not only the screens.

We hold the information needed to run your practice and let you return to it:

• Your account: the email address you signed in with, or the sign-in from Apple or Google if you used one.
• Your profile: your role, experience managing people, roughly how many people you manage, what you want to improve, and the time zone your device reports.
• Your situation: what you describe, the person you are rehearsing with, and anything you add afterwards.
• Your practice: the messages in the conversation, the feedback about it, and the notes made from it.
• Your voice during a spoken practice: it is used to make the transcript and reply. Voice recordings are not kept after that work is done.
• Whether you have Pro, and until when. It comes from the App Store rather than from you, and is kept so the app knows what you have paid for, never what you have said.

The app has no field for the nationality, gender, age, or religion of the person you are rehearsing with. It asks how they behave, because that is what you have to handle in the conversation.

It does not ask for anyone’s real name at work. If you type one, it is treated like the rest of your practice and deleted with it.

Six companies process what they need to run the app. Three of them are involved only when you speak, and one only when you pay.

• Firebase and Google Cloud store and process app data. Firestore and application functions run in Mumbai, India.
• OpenAI writes the replies and feedback. It receives the situation you described and the practice conversation so it can do that work. Your text is not used to train a model.
• LiveKit carries the live voice conversation between your phone and the app while you speak. Its service runs in Mumbai, India.
• AssemblyAI turns your voice into words while you speak. Audio is sent to them only during a spoken practice.
• Sarvam AI speaks the other person’s reply aloud. It receives the words of that reply, never your own recording.
• RevenueCat manages the Pro subscription if you buy one. It receives your account’s ID and the purchase itself, from the App Store — never anything you type or say here.

To run the practice, write feedback afterwards, and let you find a past conversation again.

Your profile changes what the app suggests and how it talks to you. We do not use it to build a picture of you for anyone else.

Your profile, situations, transcripts, feedback, and notes stay until you delete them, so you can return to an older conversation.

Voice recordings are used to turn speech into words and make the reply. They are not kept for replay in the current app.

You can delete one practice, every practice, or your whole account in Settings under “Your data”.

Deletion is immediate in the live database and removes a practice together with its feedback and notes. Encrypted backups held by our hosting provider can keep a copy for a short time afterwards, never longer than 30 days. Nothing can be restored from a backup to your account after you delete it.

You can ask Before You Say for a copy of what is held about you, for something to be corrected, or for all of it to be deleted.

You can delete data in the app without writing to us. For a copy, a correction, or help with deletion when you cannot use the app, write to saravanakumargn@gmail.com. Include the email address you used to sign in.

This app is for people who manage other people at work. It is not designed for anyone under 18.

The date at the top is when it last changed. If a change matters to how your data is handled, you are told inside the app.