Transcription Dock is a small experimental transcription utility. This notice explains the web edition's current data handling. The web edition is not a local or offline service: audio leaves your device for processing.
Uploaded audio and transcription processing
When you start a transcription, your recording is uploaded to a private Cloudflare R2 storage location under a random job identifier. The Worker validates the file and sends its audio bytes to Cloudflare Workers AI for transcription.
The source recording is deleted after successful processing. It is also deleted after an unrecoverable failure when immediate deletion succeeds. If immediate deletion cannot be confirmed, the interface reports that cleanup is pending and scheduled cleanup retries deletion. An R2 lifecycle rule provides an additional deletion backstop.
Temporary transcript results
Plain text, timestamped text, DOCX output, and limited job status information are stored temporarily in private R2 storage. Completed results expire after approximately one hour. Failed-job status is retained briefly. You can use “Delete now” to remove a completed job sooner.
There is no account system or saved transcript history. Access to each temporary job requires a capability token kept in your browser; only a hash of that token is stored with the job.
Abuse prevention and operational data
Cloudflare Turnstile is used to reduce automated abuse. Cloudflare processes the information necessary to provide and verify that service. Transcription limits use short-lived counters associated with salted hashes derived from an IP address and a browser-generated session identifier. Raw IP addresses are not written into Transcription Dock's quota records.
Content-free operational analytics record events such as page visits, transcription starts, completions, failures, audio minutes, byte counts, processing time, outcome codes, and download format. These analytics do not include audio, transcript text, original filenames, capability tokens, signed URLs, or desktop-interest email addresses.
Cloudflare may process normal request and platform information as part of operating Workers, R2, Workers AI, Turnstile, KV, and Analytics Engine. This notice does not claim that infrastructure providers never create service or security logs.
Desktop-edition interest submissions
If you submit the optional desktop-edition form, Transcription Dock stores your normalized email address, your Yes/Maybe/No response, any optional price range, and a server-generated timestamp in Cloudflare KV. A hash derived from the email is used as the record key to reduce duplicate submissions. These records do not expire automatically and are retained until they are manually removed or no longer needed for the desktop-availability notification.
The email address is collected only to notify you about Transcription Dock Desktop availability. It is not consent to an unrelated newsletter or marketing list.
Your responsibility
Only upload recordings you are permitted to process. Consider the sensitivity of a recording before using any server-based transcription service. The planned desktop edition is intended to provide a local and offline option for privacy-sensitive work, but it does not create legal or regulatory compliance by itself.
Questions or deletion requests
Email contact@transcriptiondock.com with privacy questions or to request deletion of a desktop-interest record.