Privacy Policy for DuoVox

Last updated: 9 September 2026

David Arthur Software ("DAS", "we", "our", "us") operates the DuoVox application (the "App"). This Privacy Policy explains what data the App collects, how it is used, how we protect it, and your rights. DAS is the data controller for all personal data processed in connection with the App.

Data Controller: David Arthur Software, Kirkbymoorside, York, YO62 6LJ, United Kingdom
Contact: info@duovox.net

This Privacy Policy forms part of our Terms of Use. Where David Arthur Software processes personal data, it does so in accordance with this policy and applicable data-protection law and retains its own data-protection (controller) responsibilities.

1. Information We Collect

Audio Data (Paid Tiers): When using online transcription (Standard, Professional, or Pay-As-You-Go), audio is captured from your device's microphone or internal audio and streamed in real time to third-party processing providers for transcription. Audio is processed in real time and is not stored by us. Our providers process it to return a result and may keep limited operational logs under their own privacy policies, which can change without notice — please review their policies. You are responsible for obtaining any consent required to record, transcribe, or translate a conversation.

Audio Data (Free offline mode): When a session runs offline, all audio processing occurs entirely on your device using locally installed speech-recognition models and no audio data leaves your device. If you hold Pay-As-You-Go credits without a subscription, sessions funded by those credits run online and are covered by the paragraph above.

Transcription and Translation Text: Transcribed and translated text is displayed in the App during your session. It is held in device memory and is cleared when you clear the session or close the App — unless you save it yourself. If you use the export features (for example "Export this session", or exporting subtitles from a media file), the App writes the text to files in a folder on your device; those files remain until you delete them and are not removed by the Clear button. If you turn on diagnostics, a support bundle may likewise save session text to your device. These files stay on your device unless you choose to send them to us. For paid tiers, text is sent through our service infrastructure to a processing provider for translation; it is processed in real time and is not stored by us.

Preferences and Settings: Your app settings (selected language, theme, audio source, subscription tier) are stored locally on your device and are not transmitted to any server, although your subscription tier does appear in the licensing checks and session statistics described in section 2.

Custom Vocabulary (Boost Keywords): Words and phrases you add yourself in the Boost Keywords editor are treated differently from the settings above. They are stored locally with your DuoVox settings, and they are sent to the transcription provider as part of the connection request for each online session, so that the service can recognise them — that is what they are for. They are transmitted as parameters of the request, so they may also appear in that provider’s own operational logs. Please do not enter anything in this editor that you would not want to leave your device.

Diagnostic Logs: If you enable diagnostics, log files are stored locally on your device. You may choose to share these with us for support purposes, but they are never collected automatically.

Learned Vocabulary: To improve recognition accuracy over time, the App may automatically learn individual words (never whole sentences) from your own high-confidence transcriptions and store them in a per-language list on your device. This list is limited in size and rotated automatically, is held only on your device, is never transmitted to us, and is removed when you clear the App's data or uninstall.

2. Service Infrastructure Data

The App communicates with our own service infrastructure (the DuoVox API) automatically at several points during normal use: when the App starts, to register the device and check subscription status; at intervals during an active online session or file job, to refresh the short-lived credentials used for processing, report usage minutes, and confirm the subscription is still valid; and when a session or job ends, to send a small technical summary. The App may also contact our infrastructure in response to your actions or to app events — for example when you make or restore a purchase, submit a support report, or if the App encounters an unexpected error. On paid tiers, text sent for translation is also routed through our infrastructure in real time throughout the session, as described in section 1. None of these calls include audio or conversation content, and translation text is not stored by us. To manage subscriptions, prevent abuse, and deliver the service, this infrastructure stores the following data associated with an anonymous device identifier:

• Licensing and activation: When the App starts, it contacts our service infrastructure to register the device, check the subscription status linked to its anonymous device identifier, and restore any purchases, so that the correct tier and features can be enabled. This happens on all tiers, including the Free tier. These requests contain the device identifier and app and subscription details only — never audio or conversation content.
• Subscription information: Your subscription tier, purchase verification tokens, and billing cycle dates. A purchase token is held only while your subscription is active - it is what lets us re-check with the app store that the subscription is still valid - and it is deleted as soon as the subscription ends
• Usage tracking: Minutes of AI usage consumed in the current billing cycle
• Device verification: App integrity verification results to prevent unauthorised access
• Session statistics: When a transcription session ends, the App sends a small technical summary to our infrastructure: app version, subscription tier, session length in minutes, which speech engine served the session and whether it had to fall back to a backup engine, counts of service errors, counts of captions produced, suppressed or corrected, and summary measurements of your microphone signal quality such as signal-to-noise ratio, input level and clipping. These summaries apply to all tiers, including the Free tier, and never include audio, transcriptions, translations, or any conversation content. We use them to monitor service health, plan capacity, understand overall usage of the App, and diagnose problems when you ask us for support.
• Error reports: If the App encounters an unexpected error, it sends a content-free report so that we can find and fix the defect. This can happen at any time while the App is running, and a report may also be sent on the next launch if it could not be delivered at the time. It contains only the app version, the type of error, and the location in our code where it occurred — never the error message text, your files, your settings, or any audio or conversation content. Error reports apply to all tiers, including the Free tier.

This data is stored on our cloud infrastructure provider's globally distributed network. It does not include any audio, transcription text, translation text, or conversation content. Device identifiers are randomly generated and are not linked to your personal identity by us. An identifier belongs to an installation, not to a person: installing DuoVox on a second computer creates a second, separate record, and reinstalling creates a new one. You can view your device identifier in the App under About → Contact, and quoting it is how we match a request to access or delete your data to the records we hold — which also means such a request can only reach the installations whose identifiers you are able to give us.

When a session runs in Free offline mode, captioning and translation run entirely on your device (see section 1) — no audio, transcription or translation content is sent anywhere. Note that “Free” describes the plan you are on, not necessarily how a given session runs: if you hold Pay-As-You-Go credits without a subscription, sessions funded by those credits are AI sessions and the audio and text are processed exactly as described for paid tiers until your credits are exhausted. The only data the Free tier sends to our infrastructure is the licensing checks, session statistics and error reports described above. Separately, the first time you use a language offline, the App downloads that language's model files directly from third-party model hosts (see section 3); this download requires an internet connection and sends no personal data, audio or text.

3. Third-Party Services and Sub-Processors

To deliver paid features, we use third-party processors disclosed below by category. We do not name individual providers in this policy because providers may change over time; a current list of the specific providers is available on request from info@duovox.net.

• Speech and translation processing providers: For paid tiers, audio is streamed to processing providers for transcription, and text may be sent for translation. These providers process data in real time under their own privacy policies and data-processing terms. Their policies may change without notice; we do not control and cannot guarantee provider behaviour, and we encourage you to review their policies.
• Cloud infrastructure provider: Our service infrastructure (subscription management, usage tracking, session statistics) runs on a third-party cloud platform. This provider processes data on our behalf under a data-processing agreement and may keep limited operational logs under its own policies.
• Advertising (Free tier): On Windows, the Free tier currently shows only our own in-app promotional messages; no third-party advertising network is used and no advertising cookies are set. On Android, where advertising is offered, banner and rewarded-video advertisements may be provided by an advertising network that may collect device identifiers and usage data for ad personalisation; you can manage ad personalisation in your device's settings.
• App-store billing: Subscription purchases are processed by the Microsoft Store (Windows) or Google Play (Android). Payment information is handled entirely by the relevant app store and is never accessed by DAS.
• On-device processing: The offline transcription and translation features run entirely on your device using locally installed models. No audio or text data leaves your device when using these features.
• Model training: We do not use your audio, transcripts or translations to train any model. DuoVox builds no models of its own, and the processing providers we use handle data submitted through their APIs under terms that do not permit it to be used for training their models. If that ever changes for a provider we use, we will update this policy before the change takes effect.
• Model file hosts: The models used for on-device transcription, language detection and offline translation are downloaded to your device the first time you use a language — on all tiers, including the Free tier — directly from third-party model-hosting services. These downloads are ordinary file downloads: they contain no audio, transcription, translation or personal data, though the host will see your device's IP address as it would for any download. Once downloaded, the models work without an internet connection.

4. Data We Do NOT Collect

• We do not store audio recordings, transcription text, or translation text on our own servers
• We do not collect personal information such as your name, email or phone number (unless you contact us for support)
• We do not track your location
• We do not sell your data to third parties or share it with third parties for marketing
• We do not use third-party analytics or tracking SDKs in the App; the only usage data we collect is the first-party session statistics and content-free error reports described in section 2 (and, on Android, what Free-tier advertising requires)

5. Data Storage and Security

We use appropriate technical and organisational measures to protect personal data, including encryption in transit (TLS) for all online connections, encryption of sensitive app settings stored on your device, holding only the minimum data needed to run the service, and restricting access to it. Session content (audio, transcriptions, translations) is held in device memory during your session, is saved to your device only if you choose to export it or enable diagnostics, and is never stored on our own servers. Service infrastructure data is protected by encryption in transit and at rest. No method of transmission or storage is completely secure, but we take reasonable steps to protect your information.

5a. Diagnostic Logs on Your Device

The App keeps a diagnostic log on your own device to help us investigate faults. It records what the App did — engine choices, timings, errors — and, so that caption quality can be diagnosed at all, it also records the caption text as the speech engine produced it.

One consequence is worth stating plainly: that log records the caption text before the optional profanity mask is applied. If you have switched masking on, the masked wording is what appears on screen, in the translation beside it and in every file you export — but the diagnostic log on your device may still contain the original word.

These logs stay on your machine. They reach us only if you choose to send a problem report or a support bundle, and they are not sent automatically. You can prevent a log being written at all by turning on Private Mode, which also stops any session data being kept locally.

6. Data Retention

We keep personal data only as long as necessary for the purposes set out in this policy:

• Service infrastructure / operational data (anonymous device identifier, usage minutes, device-verification data) — retained for up to 12 months after a subscription ends or is cancelled, to prevent fraud and resolve billing disputes. The one exception is a device still holding unspent Pay-As-You-Go credit: because credits never expire, that record is kept until the credit is spent. We also keep the list of purchases already credited to a device, so that a purchase cannot be credited twice.
• Session statistics (see section 2) — individual session summaries are kept on a rolling basis and automatically deleted, typically within 30 days and in any event within 12 months; after that only aggregate statistics (session counts and durations) that are not linked to any device are retained.
• Support correspondence — retained for up to 24 months.
• Sales and billing records — a record of each purchase (what was bought, when, in which currency and for how much) is kept as part of our business accounts for as long as UK company and tax law requires, currently six years. These records are needed to meet legal obligations and are not deleted on request.
• Audio and transcripts — processed in real time and not stored by us (see section 1).
• Error reports (see section 2) — retained on the same rolling basis as session statistics, typically within 30 days and in any event within 12 months.
• Settings, learned vocabulary, exported files and diagnostic logs — stored only on your device until you delete them, clear the App's data or uninstall; we hold none of it.

7. Legal Basis for Processing (GDPR)

If you are in the European Economic Area (EEA) or the United Kingdom, we process your data on the following legal bases:

• Contract performance: Processing subscription data, usage tracking, and delivering the service you have subscribed to (Article 6(1)(b) GDPR)
• Legitimate interests: Preventing fraud and abuse, ensuring service integrity, maintaining security, monitoring service health, diagnosing problems, and understanding aggregate usage through anonymous session statistics (Article 6(1)(f) GDPR)
• Consent: Displaying personalised advertisements on the Free tier where advertising is offered (currently Android only — see section 3) (Article 6(1)(a) GDPR). You may withdraw consent at any time through your device's ad-personalisation settings.
• Explicit consent: Where audio or text you choose to process may contain special-category information, we rely on your explicit consent (see section 8).

8. Special-Category Data

DuoVox is a general-purpose tool and is not intended for special-category data (such as health, legal, religious, or other sensitive details). You are responsible for any sensitive information you choose to process through the App; where such information is processed, we rely on your explicit instruction and consent. You are also responsible for obtaining any consent required to record, transcribe, or translate a conversation. DuoVox is an assistive aid and is not a substitute for a qualified human interpreter; it should not be relied on for regulated medical or legal purposes.

9. International Data Transfers

Paid tiers stream audio and text to third-party processing providers, and our service infrastructure runs on a cloud platform, whose servers may be located outside the UK/EEA (including, potentially, the United States). Where personal data is transferred outside the UK/EEA, it is subject to appropriate safeguards as required by data-protection law. We do not store this audio ourselves; providers process it to return a result and may keep limited operational logs under their own privacy policies, which can change without notice.

10. Your Rights

You have control over your data. Depending on your jurisdiction, you may have the following rights:

• Clear all session data at any time using the Clear button in the App
• Delete diagnostic logs from your device storage
• Uninstall the App to remove all locally stored data
• Request access to, correction of, or deletion of your service infrastructure data by contacting us
• Object to or request restriction of processing of your data
• Request portability of your data in a structured, machine-readable format
• Withdraw consent for ad personalisation at any time through your device settings

Because DuoVox does not store audio, transcripts, translations, or personal account data, access and deletion requests apply only to the anonymous operational data described in section 2. To exercise any of these rights, contact info@duovox.net. We will respond within one month (extendable by up to two further months for complex requests, and we will tell you if so). We may need to verify your identity first. There is no fee unless a request is manifestly unfounded or excessive. You also have the right to lodge a complaint with the UK Information Commissioner's Office (ico.org.uk) or your local data-protection authority.

11. Data Breaches

If we become aware of a personal-data breach that is likely to result in a risk to your rights and freedoms, we will notify the relevant supervisory authority — and, where required, affected individuals — without undue delay and within the timeframes required by law.

12. California Residents (CCPA/CPRA)

If you are a California resident, you have the right to know what personal information we collect, request its deletion, and opt out of any sale of personal information. We do not sell personal information. To exercise your rights, contact us using the details below.

13. Children's Privacy

The App is not directed at children under the age of 13 (or 16 in the EEA/UK). We do not knowingly collect personal information from children. If we learn that we have collected personal information from a child under the applicable age, we will take steps to delete that information as quickly as possible. If you believe we have collected information from a child, please contact us at info@duovox.net.

14. Changes to This Policy

We may update this Privacy Policy from time to time. Changes will be reflected by updating the "Last updated" date at the top of this page. For material changes, we will provide notice through the App. Continued use of the App after changes constitutes acceptance of the updated policy.

15. Contact Us

If you have questions about this Privacy Policy, wish to exercise your data rights, or have a complaint, please contact us at:

David Arthur Software
Kirkbymoorside, York, YO62 6LJ, United Kingdom
Email: info@duovox.net