The short version

PARSR records audio, transcribes it, and analyzes it with AI to build a searchable knowledge base for you. Your data is yours. We do not sell it, share it with advertisers, or use it to train AI models. You can delete everything at any time.

What we collect

Audio recordings

When you record using PARSR, the audio file is uploaded to our servers for transcription and analysis. You can also upload existing audio or video files.

Transcripts and AI summaries

The text output from transcription, along with AI-generated summaries, action items, and other analysis, is stored in your account.

Account information

Your email address and the date you created your account. We do not collect your name unless you provide it.

Usage data

Basic information about how you use the app (for example, whether a transcription succeeded or failed) to help us diagnose issues. We do not track your behavior for advertising purposes.

How we use it

To transcribe your audio and generate AI summaries and analysis
To power semantic search across your recording library
To enable the chat feature, which lets you ask questions across your recordings
To send you account-related emails (confirmation, beta access notifications)
To diagnose and fix technical problems

We do not use your data for advertising. We do not build behavioral profiles. We do not sell your data.

Third parties we work with

PARSR uses the following services to process your data. Each acts as a data processor on our behalf, meaning they process your data only to perform the service we use them for and are not permitted to use it for their own purposes.

Gladia: Transcription

Your audio is sent to Gladia to produce a transcript. Gladia's privacy policy governs their handling of audio data.

Groq: AI summaries and voice processing

Transcript text is sent to Groq, which runs Llama (an open-weight model), to generate summaries, action items, and answers to your questions. Groq's privacy policy governs their handling of text data.

Mistral: Semantic search

Transcript text is sent to Mistral to generate embeddings that power PARSR's meaning-based search. Mistral's privacy policy governs their handling of text data.

Supabase: Database and storage

All recordings, transcripts, summaries, and account data are stored in Supabase, which runs on AWS infrastructure. Data is encrypted at rest and in transit.

We do not share your data with any other third party. We do not work with advertising networks, data brokers, or analytics companies that track you across the web.

Your audio and AI training

Your recordings are never used to train AI models, ours or anyone else's. PARSR is built on open-weight models (Llama and Mistral) specifically because their weights are publicly available and their behavior can be audited. We do not have agreements in place that would allow any AI provider to train on your data, and we will not enter into such agreements.

Data retention

Free accounts: Raw audio files are retained for 30 days, then deleted. Transcripts, summaries, and embeddings are kept for as long as your account is active.

Paid accounts: Raw audio files are retained indefinitely so you can re-transcribe with improved models in the future. Transcripts and summaries are kept for as long as your account is active.

Deleted content: When you delete a recording or your account, the associated data is removed from our systems. Some data may remain in encrypted backups for up to 30 days before being permanently purged.

Your rights

Access: You can view all your recordings, transcripts, and account data within the app at any time.
Export: You can export any transcript as Markdown, PDF, or plain text.
Deletion: You can delete individual recordings or your entire account from Settings. Account deletion removes all associated data.
Correction: You can edit and correct transcripts within the app.

Security

Audio files, transcripts, embeddings, and all other user data are encrypted in transit using TLS and encrypted at rest in Supabase. Access to user data is restricted by row-level security policies, so your data is only accessible to your account.

Children

PARSR is not directed at children under 13. We do not knowingly collect personal information from children under 13. If you have sufficient evidence that an account belongs to a child under 13, contact us at [email protected] and we will investigate and take appropriate action.

Changes to this policy

If we make material changes to this policy, we will notify you by email and update the effective date at the top of this page. Continued use of PARSR after a change constitutes acceptance of the updated policy.

Contact

Questions about this policy or your data? Email us at [email protected].

Privacy Policy