The short version.
- Your saved thought library is stored on your iPhone. Spark does not currently provide its own cloud sync or cloud backup for that library.
- Voice prefers on-device transcription. If local transcription is unavailable, Spark can automatically send microphone audio to OpenAI for transcription. Temporary recovery audio can also remain on your iPhone until the capture is saved or discarded.
- Apple handles payments. RevenueCat processes pseudonymous app identifiers and subscription information to manage Spark Pro, including when checking access without a purchase.
- We do not sell your personal data or use advertising trackers. This website has no marketing analytics, advertising cookies, account registration, or mailing-list form.
1. Who is responsible
Spark is operated by Clarity App Systems, s.r.o., company identification number (IČO) 24531910, registered at Dlouhá 730/35, 110 00 Praha – Staré Město, Czech Republic. We are the controller of personal data processed to operate Spark and this website.
For privacy questions or requests, contact contact@clarityappsystems.com. Company details are available at clarityappsystems.com.
2. Your thoughts and local app data
Spark stores saved thoughts, drafts, dates, archive status, reminder information, and app preferences locally on your iPhone. Searching, organising, and selecting thoughts for resurfacing happen on the device. We do not receive your saved library simply because you write, search, or browse it.
The app uses iOS file protection for its library. Widgets use a local shared app container. If System search is enabled, thought text and related metadata are added to a local Siri and Spotlight index. Reminders and widgets may display thought text on your Home Screen or Lock Screen; control their visibility in Spark and iPhone settings.
There is no Spark login for your library, and Spark does not currently offer its own cross-device sync or cloud backup. iOS device backups may include app data under your Apple backup settings. Sharing a thought or using Apple’s system dictation, Siri, or other system services may give data to Apple or the recipients you choose, under their own terms.
3. Microphone audio and transcription
When you use Spark’s voice capture, the app requests microphone and, where needed, speech-recognition permission. It prefers Apple’s on-device transcription when a supported local model is available. If local transcription is unavailable or fails, cloud fallback can start automatically: your captured audio is sent directly from the iPhone to OpenAI for transcription. Transcription language and a small built-in vocabulary are also used to configure the service. Your existing thought library is not sent as context.
Supabase provides the authentication and security service that authorises cloud transcription. It receives a pseudonymous user/session identifier, Apple App Attest device-key information, security challenges, language and vocabulary settings, and usage counters. Network requests expose connection information, such as IP address, to the relevant providers. The Spark Supabase service does not proxy microphone audio or store your thought library.
Spark keeps a protected temporary recording and transcript on the iPhone so an unfinished capture can recover after interruption. This recovery folder is excluded from device backups. The app removes it when the capture is successfully saved or discarded; an unfinished capture can remain across restarts. Saved thoughts are text, rather than an audio-note library.
OpenAI’s API data is not used for model training by default. Its standard Realtime abuse-monitoring retention can be up to 30 days, with longer retention possible for legal or safety reasons. We do not promise zero retention or exclusively on-device processing. See OpenAI’s API data controls.
To avoid sending audio through Spark’s cloud fallback, use typed capture instead of Spark’s microphone feature. You can withdraw microphone and speech permissions in iPhone Settings. Apple’s keyboard dictation has separate Apple controls.
4. Subscriptions and purchase information
Apple processes Spark Pro payments through the App Store. We do not receive your full payment card details or Apple Account password. We use RevenueCat to validate transactions, retrieve plans, check trial eligibility, restore purchases, and determine whether Spark Pro is active.
The RevenueCat SDK creates a pseudonymous app user identifier and communicates with its service when Spark checks subscription access, even if you have not purchased. RevenueCat processes purchase and subscription history, transaction or receipt information, entitlement status, and technical context such as app/platform version, locale, and currency. These records also support subscription reporting and fraud prevention. Your thought library and microphone recordings are not sent to RevenueCat.
Manage or cancel a subscription through Apple’s Subscriptions settings. Deleting Spark or asking us to delete a service record does not cancel a subscription. Apple retains its own transaction records. See RevenueCat’s privacy information and Apple’s Privacy Policy.
5. Website visits and contacting us
This website is hosted on Vercel. Cloudflare provides domain registration and DNS. Requests to the hosting infrastructure include technical connection data such as an IP address, requested URL, request time, browser information, and response status, used to deliver and protect the site. Hosting and DNS providers may process operational and security logs under their service terms.
The site does not load third-party fonts, advertising, marketing analytics, or tracking pixels. We do not set marketing cookies or use browser storage to track visitors. Selecting a feature preview or pausing the stars only changes the current page in your browser.
If you email us, we receive your address, message, and any attachments you send. We use them to respond, resolve support issues, and handle privacy requests. Please avoid sending unnecessary sensitive information or your private thought library. Email links open your email application; the website itself has no contact or subscription form.
6. Why we process data
| Purpose | Legal basis under the GDPR, where applicable |
|---|---|
| Provide requested app features and cloud transcription; manage and restore subscriptions | Performance of the service contract, Article 6(1)(b). Cloud voice is used to provide the transcription you request; it is not needed for typed capture. |
| Secure services, prevent abuse, operate the website, and understand subscription performance | Legitimate interests, Article 6(1)(f): reliable operation, service protection, and managing the product. This does not include advertising profiles. |
| Answer support and privacy requests; preserve records required by law | Performance of a contract or legitimate interests for support, as appropriate; legal obligations under Article 6(1)(c) for statutory requests and required records. |
Device permissions control optional access to the microphone, speech recognition, and notifications. You can revoke them in iPhone Settings without losing typed capture. We do not make decisions about you with legal or similarly significant effects solely through automated processing.
8. How long data stays
- Local library and drafts: remain in the app until you delete them or remove its data. Deleting a thought removes it from the active library, but older local migration copies, device backups, and system caches may contain earlier versions. Your Apple backup settings control those backups. We cannot remotely retrieve or erase your local library.
- Voice recovery: remains locally until a capture is saved or discarded. Audio already processed by OpenAI is subject to its retention rules described above; deleting local text does not undo a prior transmission.
- Cloud voice security: challenges expire after two minutes; hourly and daily usage records are pruned during subsequent requests. Anonymous authentication and enrolled device records do not currently have automatic age-based deletion and remain until removed. Contact us to request deletion; uninstalling does not automatically delete server or Keychain records.
- Subscriptions: pseudonymous purchase and entitlement records remain while needed to operate and restore subscriptions, resolve billing disputes, prevent fraud, and meet legal obligations. They are not automatically deleted when you uninstall. Contact us for an access or deletion request. Apple may need to retain its own purchase records.
- Support and infrastructure records: retained for the time needed to resolve the request, maintain security, and meet applicable legal obligations. Hosting-provider log retention depends on the provider’s service and security requirements; this site does not maintain a separate visitor database.
We assess deletion requests against the need to identify the relevant record, any active service, and legal retention requirements. We do not promise that all third-party records disappear immediately after uninstalling.
9. Your choices and privacy rights
You can edit and delete thoughts in Spark, change notification and system-search preferences, revoke iOS permissions, and manage subscriptions with Apple. To request access, correction, erasure, restriction, portability, or to object to processing based on legitimate interests, write to contact@clarityappsystems.com. These rights apply subject to applicable law and the type of data involved. If processing is based on consent, you can withdraw it without affecting earlier lawful processing.
Because service records use pseudonymous identifiers, we may need limited information to locate your record and verify your request. We will explain what is needed rather than ask for your password or full payment details. We normally respond within one month, subject to extensions allowed by law.
You may lodge a complaint with the Czech Office for Personal Data Protection (Úřad pro ochranu osobních údajů) or your local supervisory authority. We do not sell personal data or share it for cross-context behavioural advertising.
10. Security and younger users
We use iOS data protection, encrypted network connections, device verification, and access controls appropriate to the service. No storage or transmission method is guaranteed completely secure. Avoid putting unnecessary sensitive information into a cloud transcription session, a shared thought, or a support email.
Spark is a general-purpose thought app and is not specifically directed at children. We do not knowingly solicit children’s contact details through this website. If you believe a child has provided personal data to us without the authorisation required by applicable law, contact contact@clarityappsystems.com so we can investigate and act.
11. Changes to this policy
We will update this page when Spark’s data practices change and revise the date above. If a change materially affects your privacy, we will provide an appropriate notice in the app or on this website. For questions about the current policy, contact contact@clarityappsystems.com.