Privacy Policy
What VoiceAddress processes, why, for how long, with which named providers — including our commitment that customer recordings are never used to train AI models.
Last updated: June 11, 2026
1. Who we are and what this policy covers
VoiceAddress is a software service that processes call recordings provided by business customers in order to transcribe speech, identify spoken postal addresses, validate them against mapping data, and present the results for human review.
This policy explains what data we process, why, for how long, with which providers, and what rights you have. It covers the hearloc.com website, the customer dashboard, the public API, and telephony integrations.
For any privacy question, request, or complaint, contact [email protected].
2. Our role: controller and processor
For account data — your name, email, organization, sign-in and billing records — we act as the data controller.
For the content of recordings, transcripts, and extracted addresses ("Customer Content"), our business customer is the controller and we act as their processor: we process that content only to deliver the service the customer requested, never for our own purposes.
If you were a participant in a call that a VoiceAddress customer processed (for example, you called a service company), that company is responsible for the lawfulness of the recording and is your primary point of contact for privacy rights. We will assist customers in fulfilling such requests.
3. Data we process
Customer Content: audio/video recordings submitted by customers (uploaded, sent via API, or fetched from a connected telephony provider), machine transcripts, extracted address candidates, validated addresses, confidence metadata, review decisions, and optional webhook delivery records.
Account data: name, email address, authentication identifiers and security events (via our authentication provider), organization membership, plan and billing status (via our payment processor; we never store full card numbers), API key metadata (hashed keys only), and integration configuration (telephony credentials are stored encrypted).
Operational data: request logs, IP addresses, rate-limit counters, audit events, and error diagnostics needed to run, secure, and support the service.
4. What we do NOT do with your data
We do not use Customer Content to train, fine-tune, or improve machine-learning models — neither our own nor anyone else's. Third-party AI providers are engaged through business APIs under terms that prohibit training on submitted content.
We do not sell personal data, and we do not share it for cross-context behavioral advertising ("sale" or "sharing" as defined by the CCPA/CPRA).
We do not build advertising or marketing profiles from recordings, transcripts, or addresses.
We do not listen to customer recordings except where strictly necessary to investigate a support request you initiated, a security incident, or suspected abuse — and access is logged.
5. How we use data
Customer Content is used solely to provide the requested processing: transcription, address extraction, geocoding and validation, confidence scoring, result presentation, export, and optional webhook delivery to endpoints the customer configured.
Account and operational data are used to authenticate users, enforce plan limits and rate limits, bill subscriptions, secure the service, debug failures, respond to support requests, and meet legal obligations.
Aggregated, de-identified metrics that contain no personal data (for example processing volumes and error rates) are used to operate and improve the service.
6. Data minimization and automatic redaction
Before results are stored, transcripts are automatically scanned and common payment-card, social-security-number, and email patterns are replaced with redaction markers. Word-level timing data and raw provider payloads are stripped from stored results.
Redaction is a harm-reduction measure, not a guarantee: customers must not deliberately submit regulated data categories (see the Terms of Service acceptable-use rules).
7. Retention and deletion
Stored transcript text is automatically redacted 24 hours after processing. Google-derived cached fields (geocoding and validation payloads) are removed after 30 days, in line with Google Maps Platform caching rules. Uploaded media files are deleted from processing storage once a job completes or fails.
Job records (addresses, confidence, review decisions, audit trail) are retained while the account is active so that customers keep their processing history; customers can delete individual data or the entire account dataset at any time using the export and delete controls in the product or API.
Deletion removes data from live systems promptly; copies may persist for a limited period in encrypted backups and security logs before being cycled out, and we may retain records we are legally required to keep (for example billing records).
8. Subprocessors and service providers
We use a small set of providers, each only for the function listed: AssemblyAI and Deepgram (speech-to-text), OpenAI (structured address extraction via API), Google Maps Platform (geocoding and address validation), Clerk (authentication), Stripe (payments), Twilio (only if the customer connects their own Twilio account, to receive recording webhooks and download those recordings), Cloudflare (DNS/proxy), Google Cloud (hosting), and Sentry (error monitoring with data scrubbing).
Each provider receives only the data needed for its function and is bound by its own data-protection terms. We may change providers as the product evolves; this page will be updated. Enterprise customers can request subprocessor-change notifications in a data-processing agreement.
9. International data transfers
The service is hosted in the United States, and providers may process data in the US and other countries. Where required, transfers are protected by appropriate safeguards such as standard contractual clauses offered by the respective providers.
Organizations with data-residency requirements should contact us for a written agreement before submitting production data.
10. Security
Controls include: TLS for all traffic; tenant isolation enforced on every query; a private processing backend that is not exposed to the internet; hashed-only storage of API keys; encrypted storage of integration credentials; signed webhooks (per-tenant HMAC secrets); signature verification of inbound telephony webhooks; rate limits and atomic usage quotas; audit logging; scheduled retention cleanup; and encrypted backups.
No system is perfectly secure. If we become aware of a personal-data breach affecting your data, we will notify affected customers without undue delay, consistent with applicable law, and provide the information needed for their own notification duties.
11. Your rights
Depending on your jurisdiction (for example under GDPR or CCPA/CPRA), you may have rights to access, correct, export, delete, or restrict the processing of your personal data, and to object to certain processing. Account holders can exercise export and deletion directly in the product; for anything else, email [email protected]. We respond within the timelines required by applicable law and do not discriminate against you for exercising your rights.
If you are in the EU/EEA or UK, you may also lodge a complaint with your supervisory authority. If you were a participant in a customer's call, please contact that customer first (see Section 2); we will support their response.
12. Automated processing
VoiceAddress produces automated address suggestions with confidence estimates. These outputs are advisory: the service is designed so that a human at the customer reviews results, and we do not make automated decisions that produce legal or similarly significant effects about individuals.
13. Cookies
We use only cookies and similar storage that are necessary for sign-in, session security, fraud prevention, and routing you back to the right page after authentication. We do not use advertising cookies or third-party tracking cookies on the product.
14. Children
The service is intended for business use by adults. It is not directed at children, and we do not knowingly collect personal data from children under 16. If you believe a child's data was submitted, contact us and we will delete it.
15. Changes to this policy
We may update this policy as the product, providers, or legal requirements change. The current version is always at this address with its "last updated" date; material changes for active customers will be communicated through reasonable product or email notice where practical.
16. Contact
Privacy requests, questions, and complaints: [email protected]. Please include enough detail for us to locate the relevant account or data.