Privacy Policy | Aurelia

Aurelia Digital LLC (“we”, “our”, or “us”) provides a live speech and grammar correction app that allows users to speak in their target language and receive AI-assisted corrections. We are committed to protecting your privacy and being transparent about how your data is used.

1. Information We Collect

Email and username (securely stored)
Password, if you register with email/password (stored as a secure hash)
If you sign in with Google or Apple: your name and email address as provided by those services
Speech transcriptions and AI-generated grammar corrections
Correction history, including pronunciation scores
Language preferences (app language, target language, proficiency level)
Audio recordings of what you say in conversations, kept for 90 days so you can replay them (other recordings are processed, then deleted)
Usage analytics: which screens and features you use, and app events such as completing a lesson (linked to a user ID, not your email)
Crash reports and performance diagnostics when the app encounters an error
Purchase and subscription status, if you make a purchase

We do not collect advertising identifiers or location data.

2. How We Use Your Information

We use your data to:

Provide grammar correction services
Provide pronunciation assessment and feedback
Allow access to your correction history
Let you replay your own voice from past conversations, on any device you sign in on
Authenticate your account
Process speech-to-text transcription via third-party AI services
Screen submitted content for policy violations using automated moderation
Understand how the app is used so we can improve it
Diagnose crashes and performance problems
Manage subscriptions and purchases

We do not use your data for advertising.

3. How Data Is Processed & Third-Party AI Services

When you record speech in the app, your audio is:

Transmitted securely (encrypted) to our backend servers
Sent to third-party AI services for processing (see below)
Analyzed by a third-party AI service (OpenAI) for grammar correction
Deleted after processing, unless it is a turn in a conversation. Those recordings are stored on our servers for 90 days (see Data Retention)

Third-Party AI Services

We use the following AI services to process your recordings:

OpenAI (GPT models)

Purpose: Grammar correction and feedback

Your transcribed speech is sent to OpenAI to identify grammar errors and generate corrections, explanations, and practice exercises
Text is processed and returned; OpenAI does not use API-submitted data to train their models per their API terms

Learn more: OpenAI Enterprise Privacy

Anthropic (Claude models)

Purpose: The tutor's replies in conversations

Your transcribed speech and the conversation so far are sent to Anthropic to produce the tutor's replies, reply suggestions, and conversation topics
If you have told us your name, it is included in conversations so the tutor can address you by it
Anthropic does not receive your audio recordings
Text is processed and returned; Anthropic does not use API-submitted data to train their models per their commercial terms

Learn more: Anthropic Privacy Policy

ElevenLabs

Purpose: Text-to-speech audio for word and sentence pronunciation

When you request to hear the pronunciation of a word or sentence, that text is sent to ElevenLabs to generate audio
ElevenLabs does not receive any of your personal account information or audio recordings
ElevenLabs does not use API-submitted text to train their models per their API terms

Learn more: ElevenLabs Privacy Policy

OpenAI Moderation API

Purpose: Automated content moderation

After your speech is transcribed, the text is sent to OpenAI's Moderation API to screen for prohibited content
If a violation is detected, the recording is rejected and not saved
Moderation results are not stored; text is processed in real time and discarded
OpenAI does not use this data to train their AI models (per their API terms)

Learn more: OpenAI Moderation

OpenAI Transcription

Purpose: Speech-to-text transcription

Your audio is sent to OpenAI's servers for transcription
OpenAI processes the audio and returns text transcription
OpenAI does not use your recordings to train their AI models (per their API terms)
OpenAI maintains SOC 2 Type 2 and ISO 27001 certifications
Audio is deleted after transcription is complete

Learn more: OpenAI Enterprise Privacy

Microsoft Azure Speech Services

Purpose: Pronunciation assessment and feedback

Your audio is sent to Microsoft Azure for pronunciation analysis
Azure analyzes pronunciation accuracy and provides detailed feedback
Microsoft does not use your recordings to train their AI models (per their API terms)
Microsoft maintains SOC 2, ISO 27001, and other enterprise security certifications
Audio is deleted after analysis is complete

Learn more: Azure Cognitive Services Privacy

By using the recording features, you consent to your audio being processed by OpenAI and Microsoft Azure as described above. You will be asked to explicitly accept this when you first use the recording feature.

Analytics, Diagnostics & Purchases

PostHog

Purpose: Usage analytics

We record app events (screens visited, features used, lessons completed) to understand how the app is used and improve it
Events are linked to a random user ID together with your target language and subscription tier; your email, name, recordings, and transcriptions are never sent to PostHog

Learn more: PostHog Privacy Policy

Sentry

Purpose: Crash and error reporting

If the app crashes or encounters an error, technical details (device model, OS version, app version, stack trace) are sent to Sentry so we can fix the problem
Crash reports do not include your recordings or transcriptions

Learn more: Sentry Privacy Policy

RevenueCat

Purpose: Subscription and purchase management

If you purchase a subscription, RevenueCat processes your purchase receipt from Apple or Google to manage your subscription status
RevenueCat receives a random user ID and purchase information; it does not receive your payment details, which stay with Apple or Google

Learn more: RevenueCat Privacy Policy

4. Data Sharing

We share limited data with the following third parties:

OpenAI (GPT models): Transcribed text is sent to OpenAI for grammar correction and feedback. Text is not used to train models per their API terms.
Anthropic (Claude models): Transcribed text, the conversation so far, and your name if you have given one are sent to Anthropic to produce the tutor's replies. Text is not used to train models per their commercial terms.
OpenAI (transcription): Voice recordings are sent to OpenAI for speech-to-text transcription. Recordings are not used to train AI models.
OpenAI (Moderation API): Transcribed text is sent to OpenAI for automated content moderation. Text is processed in real time and not retained.
Microsoft Azure Speech Services: Voice recordings are sent to Azure for pronunciation assessment. Recordings are not used to train AI models.
ElevenLabs: Text is sent to ElevenLabs to generate pronunciation audio when you request it. No personal account data or audio recordings are shared.
PostHog: App usage events linked to a random user ID, for analytics.
Sentry: Crash and error reports with technical device details, for debugging.
RevenueCat: Purchase receipts and subscription status, for subscription management.

We do not share your email, password, or correction history with third parties. Your name is shared only with Anthropic, as described in section 3. We do not sell your data to anyone.

5. Data Retention

Text transcriptions and corrections are stored indefinitely so users can access their history.

Conversation recordings. When you speak a turn in a conversation, we store that recording in our own database so you can play it back later from your conversation history, on any of your devices. Each recording is deleted automatically 90 days after it was made. Only you can access your recordings. We do not use them for training AI models, for analytics, or for advertising, and we do not share them with anyone beyond the processing described in section 3.

Recordings made outside a conversation, such as quick recordings, are sent to OpenAI and Microsoft Azure for processing, then deleted from our servers. Neither OpenAI nor Microsoft retains audio after processing per their API terms.

The 90-day limit applies to the audio only. The transcript and corrections from that conversation stay in your history until you delete them.

6. User Rights

You may delete your account and all associated data at any time from within the app.

Deleting a conversation also deletes its stored recordings straight away. Deleting your account deletes all of your recordings, without waiting for the 90 days to pass.

7. Children's Privacy

Our app is intended for users aged 13 and older. By using the app and creating an account, you confirm that you are at least 13 years old. We do not knowingly collect personal information from children under 13. If you are under 13, please do not create an account. If you believe we have collected data from a child under 13, please contact us at support@aureliadigital.org, and we will delete the information promptly.

8. International Users

By using our services, you consent to the processing of your information as described here.

9. Prohibited Uses and Content Moderation

Aurelia is a language learning tool. You agree not to use the app to submit content that:

Promotes hatred, discrimination, or violence against any person or group
Contains sexually explicit or graphic material
Depicts or encourages self-harm or suicide
Harasses, threatens, or exploits any individual
Violates any applicable law or regulation

All speech submissions are automatically screened by OpenAI's Moderation API before being processed. Content that violates these guidelines is rejected and not saved. Repeated or serious violations may result in account suspension.

We reserve the right to suspend or terminate accounts that misuse the service.

10. Changes to This Policy

We may update this policy from time to time. If we make material changes, we will update the effective date at the top of this page. Continued use of the app after changes are posted constitutes acceptance of the updated policy.

11. Contact Us

If you have questions about this policy, contact us at:
Email: support@aureliadigital.org
Website: https://www.aureliadigital.org