Legal
Privacy Policy
This policy describes how RazorSoft, Inc., a Delaware corporation (“PeerKit,” “we,” or “us”) handles information when you use peerkit.dev, api.peerkit.dev, media.peerkit.dev, turn.peerkit.dev, the PeerKit API, media and TURN hosts, the dashboard, and related services.
Effective date: 5 November 2026.
1. Who this covers
It covers visitors to the PeerKit website, people who create Clerk-backed PeerKit accounts, organizations that hold projects and API keys, and end users who join PeerKit sessions through our hosted pages or through customer applications (including the PeerKit Chat WordPress plugin).
If you embed PeerKit in your own product, you are responsible for your own privacy notice to your users. PeerKit processes session and account data as needed to provide the service you configured.
2. Information we collect
Account and organization
Dashboard authentication is provided by Clerk. PeerKit receives identity and organization membership needed to operate the dashboard (such as user and organization identifiers and profile fields Clerk exposes to the application). Clerk also sends its own account email, such as verification and organization invitations. PeerKit stores organization records keyed to Clerk organization identifiers.
Projects, keys, and configuration
We store project names, slugs, hashed API keys, allowed browser origins, channels, webhook endpoint configuration (including encrypted webhook secrets), and related settings you create in the dashboard or API.
Sessions and usage
For sessions you start, we store operational records such as session identifiers, status, timing, participant roles and join/disconnect times, invitation token hashes and invitation ciphertext used to re-show hosted join material, track/audit metadata for media sessions (labels, not media files), streaming-interval and usage records for plan minutes, usage events, and webhook delivery attempts (including JSON payloads). Signaling and media hosts process the network traffic required to connect participants, including IP addresses and WebRTC connection metadata. Live camera and audio may be forwarded through PeerKit media hosts while the session is active (in-memory SFU paths). TURN may relay media or data when a direct peer path is unavailable.
Messages, files, and local recordings
Text messages, binary payloads, and file bytes on a peer data channel travel between participants for the live session. PeerKit does not store conversation transcripts or those file bytes on PeerKit servers. File bytes are not sent through PeerKit signaling. The data channel is encrypted in transit. That transit claim applies to the data channel, not as an end-to-end encryption guarantee for camera or voice SFU media. Local recording, when used, stays in the viewer’s browser; PeerKit does not upload or store the recording file. Server-side recording is not a shipped PeerKit feature.
Billing
When paid checkout is enabled, Stripe processes payment details. PeerKit stores billing state needed to enforce plans (such as Stripe customer and subscription references, period bounds, Stripe event processing records, entitlement overrides, and billing audit events). Card numbers are handled by Stripe, not entered into PeerKit forms as raw PAN storage.
Contact and marketing forms
Contact submissions may be stored by PeerKit in our database (name, email, topic, optional organization name, and message) when the API contact path is used, and may also be handled by Formlyx when a published Formlyx form is embedded on the website. Availability signups on the pricing page may be stored by Formlyx while self-serve checkout is not ready.
Website analytics
Marketing pages may load CauseWise analytics (causewise.io). The hosted invitation page under /s does not load that script.
Cookies and similar technology
Clerk sets cookies required for sign-in. Your browser and network path also expose ordinary technical data (IP address, user agent) to our hosts and providers.
Logs and backups
Application and infrastructure logs may record operational events and technical metadata (such as request timing, errors, and IP addresses) used to run and secure the service. Operational database backups may be created for migrations and recovery. PeerKit does not publish a fixed public retention SLA for every log line or backup copy.
3. How we use information
- Provide, secure, and troubleshoot the PeerKit service.
- Authenticate dashboard users and authorize API keys.
- Enforce plan limits, origins, and session lifecycle.
- Process billing and respond to contact requests.
- Understand marketing-site usage where CauseWise is loaded.
- Comply with law and protect RazorSoft, customers, and end users.
4. Legal bases (EEA/UK where applicable)
Where GDPR or UK GDPR applies, we rely on contract performance (providing the service you request), legitimate interests (securing and operating the service, limited analytics on marketing pages), consent where we actually request it for a processing activity (for example, nonessential analytics where that applies), and legal obligation when we must retain or disclose information.
5. Sharing
We share information with processors who help run PeerKit, including:
- Clerk — authentication and organization membership.
- Stripe — paid subscription billing when enabled.
- Vercel — hosting for peerkit.dev.
- DigitalOcean — API, media, TURN, and related infrastructure hosts; Postgres for PeerKit data.
- Formlyx — embedded contact and availability forms where used.
- CauseWise — analytics on marketing pages.
We may disclose information if required by law or to protect rights and safety. We do not sell personal information.
6. Retention
Ending or expiring a session revokes remaining invitations for that session, clears invitation ciphertext used to re-show join links, and closes the live media room. Invitation token hashes may remain for security and auditing until the session row is purged. Application message and file payloads from the data channel are not retained as a PeerKit transcript store. Live SFU media state is held in memory for the active session and is not stored as durable media files.
PeerKit runs bounded cleanup jobs that delete or clear the following operational data:
- Ended or expired sessions, together with participants, tracks, invitations, and media-node assignments for those sessions, are hard-deleted 90 days after session end or expiration. Active sessions are not deleted. Streaming-interval and streaming-usage rows used for plan minutes are retained and detached from the deleted session identifier.
- Invitation ciphertext is cleared when an invitation is revoked, expires, or its session ends. Active, unexpired invitations keep ciphertext so hosted join material can be re-shown.
- Webhook delivery history is deleted after 30 days.
- Contact messages stored by PeerKit are deleted after 2 years, except messages marked under a support-managed legal hold. Earlier deletion may be handled through support@peerkit.dev.
- Revoked API key records are deleted 1 year after revocation. Active keys are not deleted by cleanup.
Billing, Stripe event processing, entitlement, and financial usage records are retained as needed for plan enforcement, accounting, and legal requirements. Those records are not deleted by the automated cleanup jobs above.
Application and infrastructure logs are operated toward about 30 days of retention on PeerKit hosts; exact log and backup retention also depends on hosting configuration and temporary migration backups, which are not auto-deleted by the application. Temporary database migration dumps should be removed only after their rollback window, by operators.
PeerKit does not offer self-serve full account deletion. Supported account or personal-data deletion requests are handled by support at support@peerkit.dev. Clerk-held identity data remains subject to Clerk’s controls. If you need a data processing agreement, contact us.
7. Security
We use HTTPS for PeerKit websites and APIs, short-lived participant tokens for browsers, hashed API keys at rest, and encrypted invitation material in our database. No method of transmission or storage is completely secure.
8. International transfers
PeerKit infrastructure described in our operations docs runs primarily in the United States (including NYC1 DigitalOcean resources and Vercel hosting). Providers may process data in other regions according to their own terms.
9. Your choices
- Update dashboard profile data through Clerk and the PeerKit dashboard controls available to you.
- Rotate or revoke API keys you control.
- End sessions and disable webhook endpoints you configured.
- Request access, correction, or deletion of PeerKit-held personal data where law gives you that right by emailing support@peerkit.dev.
PeerKit does not currently offer a fully automated self-serve account-wipe for every related database record. We handle supported deletion requests through support@peerkit.dev. Clerk-held identity data is also subject to Clerk’s controls and policies.
10. Children
PeerKit is directed at developers and organizations. It is not intended for children under 16, and we do not knowingly collect account data from children under 16.
11. WordPress and other embeddings
When a site uses PeerKit Chat for WordPress, that WordPress site stores the PeerKit project API key (or reads it from server configuration), registers its origin, and calls PeerKit to create or resume data sessions and invitations. The public page loads PeerKit Chat Core from the plugin package and connects browsers to PeerKit signaling with participant tokens. Message contents on the data channel are not stored by PeerKit as a transcript. Invitation email, if used, is sent by WordPress (wp_mail), not by PeerKit mail infrastructure.
12. Changes
We may update this policy by posting a new version on this page. The effective date at the top will change when we do.
13. Contact
Privacy and support: support@peerkit.dev. You may also use Contact PeerKit. Operator: RazorSoft, Inc. Company site: razorsoftinc.com. See also the Terms of Service.