Privacy Policy
This policy explains the current MVP data-handling model for uploaded call recordings and extracted address results.
Last updated: May 25, 2026
Data we process
VoiceAddress processes recorded call files that users upload, machine transcripts generated from those recordings, extracted address candidates, validation metadata, account identifiers, and operational logs needed to run the service.
Users should not upload healthcare, payment-card, social security, passport, or similarly sensitive regulated data unless they have a separate written agreement that explicitly covers that use case.
How the data is used
Uploaded recordings are used to transcribe speech, extract address mentions, validate address candidates, and present review results to the authenticated user or organization.
We do not sell uploaded recordings, transcripts, or extracted addresses. Provider APIs may process data as subprocessors to deliver transcription, extraction, geocoding, and address validation.
Retention
The current service design redacts stored transcript text after 24 hours and removes Google-derived cached validation fields after 30 days.
Users can request export and deletion from the product. Production deployments should run retention cleanup on a schedule rather than relying only on manual cleanup controls.
Security
Access to processing history is scoped to the authenticated user or organization. Administrative testing tools are protected by authentication and marked noindex.
Before broad public launch, upload malware scanning, stricter MIME verification, production rate limits, and tenant quotas should be enabled.
Your choices
Users can export or delete stored VoiceAddress job data through the product controls where enabled.
For privacy requests or business data-processing terms, contact the VoiceAddress operator before uploading regulated or high-risk data.