Last updated 6 October 2026

Privacy Policy

This describes what the Roger Android app collects, where it goes, how long it is kept and how to get rid of it. It is written from what the app actually does, not from a template.

1. Who we are

Roger is operated by Ovrthinq. For any privacy question, or to exercise any right in section 8, contact neelborad56@gmail.com.

Under India’s Digital Personal Data Protection Act 2023 we are the Data Fiduciary for the personal data described here.

2. What we collect

Account data

Your name and email address, taken from your Google account when you sign in. We do not receive your Google password.

We use your email address to send you emails about your account: a welcome when you join, and a confirmation if you delete your account. We do not send you marketing email unless you choose to receive it.

Audio you record

Roger records audio only when you start it. Recording never starts on its own. A notification stays on screen the whole time, so you always know the microphone is on.

Audio you share into Roger

Voice notes or other audio files you choose to share from another app.

Derived content

Transcripts, summaries, to-dos, topics and speaker labels produced from the above.

What we do not collect

We do not read your call logs, your contacts, your notifications or your messages. We do not track your location. Roger never records phone calls. The app declares no permission for any of these.

3. Other people in your recordings

A recording usually contains someone else’s voice, and that person is not our user.

If you are in a Roger recording and want your data removed, contact neelborad56@gmail.com and we will remove it from the account that holds it.

4. Where your data goes

Audio is encrypted on your phone before it is written to storage, and is uploaded over an encrypted connection.

We use a small number of service providers to run Roger: for hosting and storage, speech-to-text, AI summarisation, push notifications, and email. Each is engaged under contract to process data only on our instructions. Some of them process data outside India.

The current list, including what each provider does and where it operates, is at roger.ovrthinq.com/subprocessors. We keep that page up to date and will tell you in the app before a change that materially affects how your data is handled.

Because of this, your audio and transcripts are processed outside India in some circumstances. If that is not acceptable to you, please do not use Roger. The primary speech-to-text path runs in India.

Summaries, to-dos and Ask answers are produced by a large language model, reached through the AI provider listed on our sub-processors page, on paid endpoints, not free tiers. The text sent is your transcript. We do not use your recordings or transcripts to train our own models, and we do not sell them or share them for advertising. We are reviewing our AI provider’s retention and training terms and will state their position here once we have confirmed it, rather than promise something we have not verified.

We do not sell your data. We do not use it for advertising.

5. What we use it for

Only to provide the product: transcribing your recordings, writing summaries and to-dos, letting you search and ask questions across your own memories, and recognising voices you have named so future recordings are labelled.

We do not use your recordings or transcripts to train our own models.

Sharing

You can create a link to a memory to show it to someone. A share link is a public web page: anyone who has the link can open it without a Roger account. The page is not indexed by search engines, but it is not secret either, treat the link itself as the key.

If you share a memory by email with someone who does not use Roger, we send them one email on your behalf, with your name, the memory’s title, any message you add, and the link. Nothing else from the memory goes into the email itself.

You choose how much it carries, and the choice is shown to you before you send it:

A share is a snapshot, not a live window: it shows what you chose at the scope you chose, and editing or deleting the memory afterwards does not widen it. We do rebuild that snapshot when the same memory is corrected, for example when Roger finishes working out who was speaking, or when you name a voice, so a link you already sent shows the corrected version. It is always rebuilt at the scope you originally chose, and never a wider one.

Any link can be revoked, which takes the page down immediately and permanently. The in-app control for this is still being built: until it ships, email neelborad56@gmail.com and we will revoke it for you. Deleting the memory also destroys every share of it.

6. How long we keep it

Audio is deleted after 30 days by default. You can change this in Settings → Delete audio after, including to 0 days, which deletes audio as soon as it has been transcribed.

Transcripts and summaries stay until you delete them.

Deleting a memory deletes its audio, transcript, summary and to-dos. Deleting your account deletes everything, including your voice signatures, see how to delete your data.

7. Security

No system is perfectly secure, and we will not claim otherwise.

8. Your rights

Under the DPDP Act 2023 you may:

Account deletion is immediate and cannot be undone.

9. Children

Roger is not intended for anyone under 18, and we do not knowingly collect data from children.

10. Changes

If this policy changes materially we will tell you in the app before the change takes effect.

← Back to Roger