Offline development update · 14 October 2026
The latest development app generates Gemma 4 E2B minutes on the device after an explicit model import, and encodes MP3 locally in project and batch exports. These operations do not upload audio or transcripts or switch to a cloud service. Offline speech also supports vocabulary hints. The latest source now includes local anonymous speaker suggestions for transcribed recordings up to three minutes, using a bundled, verified model. Suggested labels require review and can be undone; mixed or overlapping passages remain unassigned. No reusable voice profiles or real-person identification are provided. The imported Gemma model can be removed without deleting projects and is excluded from device backup; bundled speaker/separation models are removed with the app; opening an external model download link is a separate network action.
The latest development app also separates voice/vocals and background on the device using a bundled verified Spleeter model, without uploading a recording or using processing credits. It previews, balances and exports real stems as AAC/M4A, MP3, WAV, FLAC and AIFF; approved redactions are applied before processing and to both saved stems. The private original is preserved. Tests cover selected recordings and simulators, not guaranteed quality: separation may remove wanted speech, leave leakage or introduce artifacts, and vocal bandwidth is limited. It cannot reliably isolate each overlapping speaker. Physical-device and broader listening checks remain open.
The latest development source asks questions across user-selected transcripts locally with Gemma and displays exact cited quotations for review. Private saved question checkpoints contain selected quotations and project references, are excluded from backup and bounded to32MiB, and can be cleared in Storage. Deleting a project discards these disposable checkpoints. Removing the model preserves saved checkpoints, so complete unchanged answers can still reopen without inference. Offline search/questions are included in Local Tools or Pro after model import. Phone/iPad saved-answer replay and original source playback pass real simulator UI checks; Phone missing-answer replay and real-request cancellation/relaunch checks also pass with no invented citation. Fresh-inference UI, interruption at every stage, natural-recording accuracy and physical performance remain release gates. current cross-meeting continuity still has a legacy explicitly approved service path. Stem exports now support verified AAC/M4A, MP3, WAV, FLAC and AIFF settings; converting a lossy recording to a lossless file does not restore lost quality.
Prepared exports in the latest development source recheck the complete original file and reviewed edits before sharing, using incremental SHA256 reads. Simulator checks verify an approved privacy AAC/SRT copy, complete audio decoding, preserved original bytes, and actual local Save to Files on iPhone and iPad with matching destination bytes. The export-ready screen also offers a direct full-screen Files picker for accessible iPad use. This does not guarantee removal of every sensitive paraphrase or compatibility with every destination; review share copies before sending them.
Simulator and real-file checks are documented; they do not establish physical-device performance. The uploaded Build 21 predates these changes and still has the legacy approved service paths described below. Full offline conversion, AI accuracy, summary length and long-recording support, and physical-device verification remain incomplete. SoundBrief is not yet released to clients.
SoundBrief · Privacy Policy
Chinese script copies · offline development tool
Project tools → Translate transcript → Chinese script creates a separate Simplified or Traditional copy on iOS 17 or later, using native character conversion. No account, AI model, language download or recording upload is needed. This changes characters; it does not translate Cantonese into Mandarin or correct a transcription. Review names and ambiguous characters. The private original, speaker labels and source passage times are preserved.
Saved copies stay in the private, backup-excluded translation history. Delete a chosen copy or its project to remove it. Shared SRT, WebVTT, TXT and JSON copies exclude approved protected passages and recheck the original and current edits before export. Files you deliberately save or share follow the destination’s retention settings. Phone and iPad simulator checks verify conversion, relaunch recovery, actual Files saves and unchanged original bytes; physical-device checks remain open. Apple language translation is a separate iOS 18 or later operation with device-specific availability. The uploaded Build 21 predates this feature; the app is not released.
Speech activity · offline development tool
Project tools → Find speech locates suggested speech regions using a small model included in the app. No account, model download or recording upload is needed. Selecting a region selects the corresponding retained audio; it does not automatically cut sound or infer words or speaker identities. Quiet speech, overlap and singing can be misclassified, so listen before editing.
Saved activity timings stay in private app storage, excluded from backup, and can be cleared in the tool or removed with the project. Temporary analysis audio is deleted after completion or cancellation; interrupted owned working files are recovered on the next analysis. Originals are preserved. Phone and iPad simulator checks cover real inference, replay and selection; physical performance, manually labelled boundary accuracy and long-file stress remain release checks.
Local recording, extraction and editing do not require an account. Online content processing requires your approval. SoundBrief contains no advertising SDK, tracking SDK or content analytics. This policy describes the current development build; public production online processing is not yet available.
Operator and local information
SoundBrief is provided by MW Supply Limited. The app stores copied original media, extracted audio, edits, transcripts, speaker names you enter, summaries, project organisation, settings and recovery history on your device. Microphone permission is needed only for recording. System pickers let you choose media from Photos or Files, including iCloud Drive. Selecting a file is not approval to upload it to SoundBrief's service.
Private transcription uses an available Apple on-device recognizer and stops if the requested local capability is unavailable. Device backups and files you save to iCloud Drive are governed by your Apple settings. SoundBrief does not offer automatic cloud project sync in the current build.
Optional online processing
After approval, supported operations send selected audio or transcript content, vocabulary hints, speaker labels, source timestamps, questions and processing options to the configured service. The service processes these to return transcripts, minutes, answers, separated stems or exports. Account identifiers, request identifiers and usage records associate jobs with the correct account, enforce allowances and prevent duplicate charges. Transcripts and recordings are not put into analytics or application logs.
The current evaluated service uses self-hosted model weights. It does not use your recordings or transcripts to train models or target advertising. Public hosting, region and operational retention configuration must be finalised before production activation. The app checks service disclosures before sending approved content. A future processor or data-use change requires updated disclosures and any applicable consent; this page does not authorise an undisclosed provider.
Apple services
Optional Sign in with Apple supplies an account identifier and authentication credentials; SoundBrief does not request your name or email scope. Apple handles payments. SoundBrief uses verified purchase and subscription records for access, restoration, allowance delivery and refund or revocation reconciliation. We do not receive payment-card details.
If you choose song recognition, SoundBrief sends an audio fingerprint to Apple's Shazam service after its network disclosure. Recognition is distinct from transcription and may return no match. Optional translation uses Apple's supported translation framework and may require a language-model download. Apple services are subject to Apple's privacy policy.
Retention and deletion
Local projects remain until you delete them. They include original media and edit history. Generated temporary share copies can be removed in Settings → Storage after closing the share sheet; abandoned temporary exports are also cleared on a later app launch. Copies already saved or sent elsewhere, and system backups, must be managed separately.
The current test service defaults to 24-hour retention for uploaded content, processing jobs and results, including recoverable summary sections. That is a test configuration, not a guarantee about a future production host or its backups. Production retention and processor details will be published before that service is enabled.
For a configured account, Settings → account controls → Delete cloud account revokes access and requests cancellation and removal of server content. Check the deletion status: accepted and completed are different states, and external cleanup may require retries. Minimal accounting records may remain to reconcile purchases and prevent allowance resets. Local projects remain until separately deleted. Apple subscriptions must be cancelled separately in your Apple account settings.
For access or deletion requests, or help with a pending deletion, contact us below. We may need enough information to verify the request. We will explain any information we must retain and why rather than promise unverified erasure.
Speaker labels and privacy copies
Speaker labels distinguish speech within recordings; they do not identify people by voice. Names are entered manually. No reusable voice profile is created by default.
Sensitive-passage suggestions run locally and require review. Approved redactions apply to supported share-copy exports while the original stays private. Suggestions are not comprehensive: review repeated mentions, contextual clues, audio and generated text before sharing. Recipient services handle shared files under their own policies.
Support and this website
If you email support, we receive your email address and the message or attachments you choose to send. We use them to answer and investigate your request and retain correspondence as needed for that purpose and related support history. You may ask for deletion. Please avoid sending sensitive recordings or transcripts.
GitHub Pages hosts this website and may process technical requests, including IP addresses, to operate and secure it under the GitHub privacy statement. These SoundBrief pages contain no analytics scripts, advertising, tracking pixels or contact form.
Changes and contact
We update this policy and the corresponding App Privacy disclosures when data practices change. The date above identifies this version. Contact MW Supply Limited at mwsupplylimited@gmail.com.