Privacy Policy

The short version

• Your audio is never stored. Speech is transcribed in-flight and discarded.
• Transcripts live in your history for at most 30 days, or until you delete them.
• We never use your content to train models.
• Edit text is shared with us only if you switch on “Help improve SetuInk” — off by default.
• Export or delete everything, instantly, from your dashboard.

What we process

Audio streams from your microphone to Amazon Transcribe (AWS, us-east-1) over an encrypted connection and is processed in real time. Neither SetuInk nor AWS retains the audio. In demo mode, speech recognition happens inside your own browser and never reaches our servers at all.

Transcripts are sent to Amazon Bedrock (Nova models) for cleanup and analysis, stored in your dictation history (DynamoDB, encrypted at rest) with a 30-day automatic expiry, and shown only to you.

Account data: a user identifier, optional email (if you sign in), your settings, dictionary terms, snippets, and word-usage counters.

Quality telemetry: per-request metadata (which writing profile and language you used, which AI model ran, latency, your ratings, and how much you edited the output — as a category, not the text). The text of your edits is included only when you have opted in.

What we never do

• Store or listen to your audio.
• Train models on your content. AWS Bedrock does not train on customer content either.
• Sell or share your data with advertisers or data brokers.
• Route your voice through third-party subprocessors beyond AWS — with one exception, described immediately below.

The one exception: browser-based recognition

Eight regional Indian languages — Tamil, Telugu, Kannada, Malayalam, Bengali, Marathi, Gujarati and Punjabi — are not yet available on our own transcription service. When you choose one of them, SetuInk uses your browser's built-in speech recognition instead. In Chrome and Edge, that means your audio is processed by Google rather than by us, under Google's terms and not ours. In Safari it is handled by Apple.

We mark these languages in the language picker so the choice is visible at the moment you make it. Every other language uses our own transcription, in a single cloud region, with no third-party processor. If this matters to you, pick a language that is not marked, or contact us before dictating sensitive material.

Cookies, IP addresses and logs

SetuInk sets three first-party cookies, all strictly necessary: a signed session cookie once you sign in, an anonymous identifier so guest dictations and usage limits work before you have an account, and a short-lived sign-in state cookie that protects the login round-trip. We set no advertising or analytics cookies, and we run no third-party analytics or tracking scripts, so no consent banner is required.

Our content delivery layer processes your IP address to route requests and to enforce rate limits and abuse protection. A one-way hash of your IP is used to apply free-tier limits to guest users. Server logs record request metadata and error types; they do not contain your transcripts or your audio. Logs are retained for a limited operational period and then deleted.

Your rights (incl. DPDP)

Under India's Digital Personal Data Protection Act and similar laws elsewhere, you can access, correct, export, and erase your personal data. In SetuInk these are self-serve: Dashboard → Your data → Export gives you a complete JSON copy; Delete account data erases history, dictionary, snippets, and usage records immediately — no email required, no waiting.

Where data lives

All processing and storage happens in Amazon Web Services, region us-east-1, encrypted in transit (TLS) and at rest. API keys are stored only as SHA-256 hashes.

Grievance Officer

If you have a complaint about how your personal data is handled, contact our Grievance Officer at privacy@setuink.com with “Grievance” in the subject line. We acknowledge within 72 hours and aim to resolve within 30 days, as required by India's Digital Personal Data Protection Act.

If your grievance is not resolved to your satisfaction, you may escalate it to the Data Protection Board of India.

Consent, and withdrawing it

We process your data to provide dictation: to transcribe what you say, to polish it, and to keep the history and dictionary you ask us to keep. You give consent by using the service, and you can withdraw it at any time — withdrawing is as easy as giving it. Dashboard → Your data → Delete erases your stored data and ends processing; you can also stop by simply not dictating. Withdrawal does not affect processing already carried out.

SetuInk is not directed at children. We do not knowingly collect personal data from anyone under 18, and we do not profile or serve targeted advertising to anyone. If you believe a child has used the service, contact us and we will delete the account.

Where your data is processed

Our servers and databases run in a single cloud region in the United States (us-east-1). If you are in India or elsewhere outside the US, your data is transferred there to provide the service. Content delivery and TLS termination happen at edge locations closer to you; those handle requests in transit and do not store your transcripts or audio.

If something goes wrong

In the event of a personal-data breach, we will notify the Data Protection Board of India and affected users without undue delay, describing what happened, what data was involved, and what we are doing about it.

Contact

For data-rights requests (export, deletion), email privacy@setuink.com — or use the self-serve controls in your app dashboard.