Protecting your business records
Reviewed 14 October 2026. This page describes implemented controls and recorded verification. It is not an independent certification or a guarantee of uninterrupted service.
Access and tenant boundaries
Operator requests use signed sessions and verify current company membership, role and access on the server. Crew sessions are company scoped. Owner and administrator permissions protect financial settings and evidence sharing. Account security supports authenticator verification. Private client evidence uses revocable, expiring links with selected records only; anyone holding a link can read its authorised contents until expiry or revocation.
Files, transport and application protection
HTTPS and protective browser headers secure transport and reduce browser attack exposure. Uploads are checked for supported type, size and content. Private malware scanning quarantines new files until a clean result. Content hashes and storage-generation checks prevent a scan or sharing permission from approving replacement content. Access to uploaded files uses short-lived signed URLs after authorisation. Secret scanning and regression checks run in the release workflow.
Hosting and overseas processing
The primary Firestore database and uploaded-file storage are in the United States. Vercel processes the website and application on its global infrastructure. AWS email delivery uses Sydney. See the Privacy Policy for processing and personal-data practices. We do not claim Australian-only hosting.
Backup and recovery evidence
Database protection includes point-in-time recovery, daily backups retained for 35 days and deletion protection. File versioning and a seven-day soft-delete window provide additional protection. Encrypted recovery packages cover Authentication, configuration and file copies.
A timed isolated recovery exercise on 14 October 2026 completed in 179 seconds. It compared a database snapshot, restored Authentication to a localhost emulator, verified file checksums and decrypted the recovery package. Production customer records were not overwritten. This exercise demonstrates that tested snapshot and scope; its duration is not a promised recovery time for every incident.
Procedures require a timed recovery exercise after material identity or storage changes and at least quarterly. Recovery objectives and contractual service levels require a separate written agreement.
Monitoring and incident response
Multi-region uptime checks monitor the website, login and API health. Health checks cover scheduler freshness, backup freshness and prolonged scanning backlog. Sanitised operational errors and owner alerts support investigation. Incident procedures cover impact assessment, evidence preservation, credential replacement, regression rollback using current secrets, recovery verification and factual customer communication.
Report suspected security issues privately to support@trafwise.com. Include the affected URL, time and a safe description. Do not send passwords, private keys or customer documents. Support is prioritised by impact; this page does not promise staffed 24-hour support.
Your records and account closure
Owners can request a company export including newer equipment collections and nested records. Privacy requests follow a reviewed identity and retention process. Signed payroll and safety evidence may require lawful retention; cancellation does not automatically erase employer business records. See data requests, privacy and terms.
Assurance boundaries
TrafWise does not claim ISO 27001 or SOC 2 certification, independent penetration-test accreditation, zero defects or unlimited capacity. Existing company roles remain the access boundary; depot grouping does not provide branch-level isolation. We describe tested controls and their limits so customers can assess the service accurately.