Who handles your data
Ark.in, the operator of Pristi, is responsible for deciding why and how personal data is processed. For privacy questions, consent withdrawal, rights requests, or complaints, contact [email protected]. You may write in English or Hindi.
Our privacy principles
- We do not sell personal data.
- Voice input is used only to generate the text you asked for, not to build an advertising profile.
- Pristi's backend does not intentionally store audio or transcripts from normal processing.
- Pristi never captures or reads your screen. There is no context-awareness of any kind — this is a deliberate contrast with tools that screenshot your screen to guess what you're doing. Pristi only hears what you say after you press a hotkey.
Data we handle
What leaves your device:
- Account data, such as your email address, sign-in verification, account identifier, and authentication tokens, is used to create and protect your Pristi account, connect paid access to the correct user, and maintain account continuity.
- Audio you record (
Alt+X for Work mode, Alt+C for Chat mode) is sent to Pristi's server to be transcribed and cleaned, then returned as text. This is the core function of the product and cannot be turned off.
- Transcript, generated output, and writing settings, such as mode and polish level, are processed to create the writing result you requested. Pristi's backend does not intentionally retain this content after normal processing.
- Non-content diagnostics are sent automatically for every dictation: app and Windows version, mode, polish level, plan tier, timings, and error codes. Never your audio, transcript, or output.
- If you purchase Pristi Pro, the payment details you enter go directly to the payment provider. Pristi receives only the limited purchase, subscription, account-reference, and entitlement information needed to activate and support paid access. We do not receive your full card number, UPI ID, or full payment credentials.
What Pristi does not collect: your screen contents or unrelated clipboard contents. Your transcript and finished text are processed only to provide the requested result and are not intentionally stored by Pristi's backend after normal processing.
Stored locally on your machine: your settings and sign-in tokens, and audio written to disk briefly before each request. That audio is deleted as soon as the dictation resolves; only a genuine failure leaves it on disk, and anything older than 7 days is deleted automatically.
A note on the clipboard: Pristi delivers dictated text via your clipboard. If you have Windows Clipboard History turned on (Win+V), Windows itself stores everything placed on the clipboard, including every Pristi dictation — this is Windows behavior, not something Pristi can control or exempt itself from. If you dictate sensitive text, turn off Clipboard History in Windows Settings → System → Clipboard.
Diagnostics and product improvement
Pristi sends non-content diagnostics by default so we can keep the service reliable: processing time, failure stage, delivery path, entitlement tier, app version, Windows version, and recording duration. These are not intended to include your audio, transcript, generated text, clipboard contents, or screen contents.
Service providers and sharing
Pristi shares data only with categories of service providers needed to deliver and protect the service:
- Speech and language processing providers: process recorded audio, transcripts, and generated text to produce the dictation result you request.
- Cloud infrastructure providers: provide hosting, databases, security, and service monitoring.
- Account authentication providers: verify sign-in and help protect your account.
- Payment providers: process payments and send Pristi the limited purchase status needed to grant and support paid access. Pristi does not receive your full card number, UPI ID, or full payment credentials.
These providers may process data in India or other countries where they operate, subject to applicable safeguards and contractual controls. We do not sell personal data or share it for advertising.
Retention
- Audio: not intentionally retained from normal processing; temporary local files and backend buffers are cleared after processing. A failed dictation's audio is kept locally on your machine for at most 7 days.
- Transcripts and generated output: not intentionally stored by the backend from normal processing.
- Diagnostics: retained for reliability analysis, typically up to 180 days.
- Payment and entitlement records: the payment provider retains payment records under its own policy. Pristi retains the limited purchase, subscription, account-reference, and entitlement records needed for billing support, fraud prevention, legal compliance, and access management.
Your rights, choices, and consent
- You can sign out at any time from Settings.
- You can request account and data deletion from the data deletion page — the same request form and process used for Pristi's Android app, since accounts are shared across both.
- You can contact us for access, correction, deletion, or billing questions.
- You may withdraw consent by stopping the relevant optional feature, signing out, or submitting a deletion request. If processing is necessary to provide a feature you requested, withdrawing consent may prevent that feature from working.
- If a privacy complaint is not resolved, you may use any remedy available under applicable Indian law, including a complaint to the Data Protection Board of India where available.
Children
Pristi is not directed to children and is not intended for users under 18 unless a parent or legal guardian has approved use of the service.
Changes to this policy
We may update this Privacy Policy as Pristi Desktop changes. If we make material changes, we will update this page and the date above. The privacy policy URL should remain stable even if the website design changes.
