Last updated: October 21, 2026
For questions or requests about your data, contact us through the app’s support chat (“Contact support” in Settings), or email [email protected].
For practical help with Oratio, read Oratio Help and Support.
Oratio is published by Florian Bidaut, the controller of your personal data, at 145 bis rue de Silly, 92100 Boulogne-Billancourt, France.
Oratio collects information you provide, information the app produces when you use it, and technical information sent automatically.
Recordings and transcripts. Oratio records your voice using the microphone. Your recording is transcribed and analyzed by AI only if you have allowed it.
Results and history. Your recordings, transcripts, scores and feedback are stored on your device. If you have an account and enable backup, the results, transcripts and topics of your speaking sessions are also copied to your account. Results from targeted exercises stay on your device. Your audio recordings are never backed up to your account.
Getting-started questionnaire. On first launch, Oratio asks for your first name, age range and gender, which are required to continue, as well as your goals, confidence when speaking, practice schedule and how you heard about the app. In versions offering a choice of practice language, Oratio also asks whether the selected language is your native language. These answers personalize your practice. Your first name stays on your device and, if you have an account, in your account. The other answers are also used for usage analytics.
Account. An account is optional and is created with “Continue with Apple”. Apple provides an identifier specific to Oratio and, on your first sign-in, your name and email address. Apple may hide your email address using a relay address. This address is stored with your account and is not shared with other providers.
Purchases and subscription. Apple handles payments for and management of Oratio Pro. Oratio never receives your payment card or bank details. Oratio and its subscription-management provider receive your subscription status and purchase history. This provider is contacted at each launch, even if you do not have a subscription, with a subscriber identifier and information about your device.
Usage and technical data. Usage analytics are enabled by default. They may receive screens and actions, questionnaire answers other than your first name, the speaking format and topic category, the daily challenge associated with a take, your scores, the category and number of detected filler words, technical measurements of your takes such as pauses and detected language, errors, your plan and total paid, subscription events and analysis costs. This information is associated with a random installation identifier, or your account identifier, device information and the country and region inferred from your IP address. Usage analytics never receive your recording, transcript, feedback, first name or email address. Technical logs also retain information about analyses and errors.
Notifications. If you have an account and allow reminders, Oratio stores your device’s notification token, language and time zone.
Support chat. If you contact support in the app, your messages and attachments are sent to our customer-support provider. Oratio includes your first name, installation or account identifier, plan, whether you have an account, app version, iPhone model and iOS version. A random support-specific identifier stays on your phone so you can return to your conversation. Your email address is not sent unless you include it yourself. To notify you about replies, the app also provides this service with your device’s notification token at each launch.
Your voice is sent to an AI service only after your explicit permission, requested after the getting-started questionnaire or when starting a take. This permission covers transcription after recording and, in versions offering Filler Hunt, the live transcription described below. Once you have given this general permission under these conditions, the exercise does not request separate permission. Without this permission, audio transfers to AI services are blocked; the rest of the app remains accessible.
When you finish a speaking take: