Data & Privacy
How Your Data is Handled in welove
We only collect information necessary for authentication, learning progress, and chosen features. Here is our data flow explained clearly.
What Data We Process
- Account details: email address, cryptographic password hash, and display name if provided.
- Guest and standard sessions: technical authorization tokens used to maintain access.
- Learning state: selected exercises, submitted responses, accuracy scores, SRS review intervals, and course progress.
- If notifications are enabled: selected topic preferences, local notification schedule, and encrypted device push tokens. Notifications are delivered by your browser's or device's push service (for example Google Firebase Cloud Messaging, Mozilla, or Apple), which receives only the delivery address and the notification text.
- Aggregated technical analytics: action events, page routes, and system telemetry — never including email addresses, study answers, or audio recordings.
- For user-submitted materials: provided URLs and associated study data.
Speech & Voice Practice
Voice recordings initially remain in local browser memory. Audio is sent to our server only after explicit user consent and clicking 'Transcribe'. Audio is processed strictly in-memory: raw audio recordings are never stored in databases, log files, telemetry, or crash reports.
Transcription runs on our infrastructure by default. If fallback mode is enabled, audio may be securely routed to Groq (US) solely for transcription after an on-screen notice. Transcribed text can be reviewed and edited prior to submission.
Speech practice is entirely optional; all tasks can be completed in text mode. Before transmission, we store consent verification (policy version, scope, and timestamp only), with zero voice audio or answer content recorded.
Translation, AI & Audio Synthesis
For select educational features, text prompts may be processed by AI services (OpenAI, Gemini, or compatible providers). Subtitles of learning videos are translated and checked via DeepSeek or a compatible service such as OpenRouter. Text-to-speech audio for lemmas and phrases uses OpenAI or ElevenLabs. These requests never include passwords, auth tokens, or user voice recordings.
Generated TTS audio may be temporarily cached for 24 hours to prevent redundant external API calls. This is synthesized voice audio, never user microphone input. Expired cache files are deleted automatically.
Data processing regions depend on the respective third-party provider policies rather than welove infrastructure.
Why We Use Your Data
- To authenticate your account and synchronize progress across sessions.
- To generate your daily learning route and calculate spaced repetition schedules.
- To provide customer support, resolve errors, and protect the service from abuse.
- To improve product experience using aggregated technical metrics without exposing private answers or voice.
Data Retention, Cookies & Account Deletion
Authentication tokens are stored in the browser solely to maintain session state. We do not use advertising tracking cookies. Temporary audio recordings live in browser memory and are discarded upon reset or page exit.
Learning data is retained while your account is active. In account settings, you can export a full data archive or trigger permanent account deletion (cancels scheduled notifications, removes device tokens, and revokes active sessions).
Google Analytics (GA4) does not load or transmit events without explicit consent. Your choice is stored in localStorage without advertising cookies. YouTube videos are embedded with the YouTube player, and Google's privacy terms apply when an embedded video loads or plays; the main library video player uses privacy-enhanced mode (youtube-nocookie.com).
Requests & Inquiries
For data access, correction, or deletion inquiries, contact [email protected]. Do not send passwords, credentials, or raw audio files in support emails.
Third-Party Media & Age Requirements
When submitting URLs or learning materials, ensure you have the right to use them. welove does not grant rights to third-party videos, subtitles, thumbnails, or translations. Public pages only include welove original content; imported video transcripts and derivative assets remain restricted within the app.
The service is intended for learners aged 18 and older located in Kazakhstan. Users under 18 or outside supported jurisdictions should not create accounts until dedicated announcements.
Last updated: July 15, 2026.