Privacy Policy

1. Overview

VeloxWaves is a privacy-first desktop AI assistant. The core features - local dictation, push-to-talk transcription, searchable second brain memory, document translation, and local AI replies - are designed to run on your computer. We collect only the account, billing, release, usage-limit, and support information needed to operate the service.

For a plain-language explanation of those data flows, read our Trust & Privacy guide. It summarizes the product architecture and does not replace this policy.

2. Information We Collect

• Account email, authentication identifiers, and basic OAuth profile information.
• Plan status, trial start and end dates, subscription status, and Stripe customer identifiers.
• Local transcription usage events needed to enforce Free plan limits, such as word counts and timestamps, not transcript content.
• Release and updater metadata, such as app version, target platform, and installer request details.
• Support issue details, system information, logs, screenshots, or attachments only when you choose to submit them.

3. Voice, Transcripts, and Local AI Data

Voice audio is processed locally by the VeloxWaves desktop app. We do not upload voice audio for transcription, do not store voice recordings on our servers, and do not store your transcript content on our servers.

Your transcription history, knowledge graph, entities, relationships, embeddings, local AI prompts, generated Brainstorm replies, and document translation outputs are stored locally on your device unless you explicitly choose to include something in a support request or enable a future cloud feature with a clear opt-in.

4. Second Brain and Knowledge Graph

VeloxWaves can build a private second brain from what you dictate. Search indexes, semantic embeddings, graph nodes, graph edges, projects, imported document chunks, and conversation memory are stored in local databases on your machine. This memory is not uploaded to Neural Wave Labs servers for hosted processing.

5. Translation, Documents, and Model Downloads

Text and document translation are designed to run locally after the required model is available on your device. Imported documents, extracted blocks, OCR results when enabled, and successful translation records are stored locally.

Some local AI capabilities may require an explicit model download. Download requests contact model or release hosting endpoints and may include basic request metadata, but the downloaded model does not grant us access to your voice, transcripts, documents, translations, or prompts.

6. Payments

Payments are handled by Stripe. We store Stripe customer and subscription identifiers so we can manage Trial, Free, Pro, and Enterprise plan status. We do not see or store card numbers, bank details, or full payment credentials.

7. Third Parties

• Supabase: Authentication, account records, plan status, trial state, and limited usage-limit metadata.
• Stripe: Payment processing, subscription state, invoices, and billing portal.
• Vercel: Website, API routes, release metadata, updater endpoints, and support submission hosting.
• Model and release hosting: Installer downloads, updates, and explicit local AI model downloads.

8. Telemetry and Support

VeloxWaves does not use hosted transcription telemetry to read your content. Support submissions may include logs or system details only when you choose to send them. We use support data to diagnose issues, improve reliability, prevent abuse, and respond to your request.

9. Data Retention and Deletion

Account and billing records are retained while your account is active or as required for legal, security, and accounting obligations. Support attachments are retained only as needed to resolve the issue. Local transcripts, memory graph data, translations, downloaded models, and document caches remain on your device and can be deleted from the app or by removing local app data. You may request deletion of server-side account data by contacting us.

10. Contact

Privacy questions can be sent to contact@veloxwaves.com.