Security

Security is where Vericia begins

You hold the identity documents, ownership records and bank details of people who have placed their trust in you. Vericia is designed, developed and tested so that only the right people in your firm can see them, and every access is recorded.

OWASP ASVS 5.0 Level 2OWASP Top 10:2025NIST SSDFNIST SP 800-63B-4Aligned with CISA guidance

In the product

Access that mirrors your organisation

Access is denied unless a role expressly grants it. Roles combine modules, actions and jurisdictions, so a Jersey administrator never sees a Mauritius file.

Single sign-on with Microsoft Entra ID

There are no separate passwords to manage. Your multi-factor authentication, conditional access policies and leaver process apply to Vericia from day one.

Least-privilege access

Read, edit and approve rights are granted by module and by jurisdiction through teams, each limited to the entities it looks after. Permissions are verified on every record, not just every screen.

Strict data segregation

Every record belongs to a single firm, identified from the signed-in account and never from the request itself. One firm's data can never be reached by another.

Data masking

Tax numbers, identity documents and bank accounts show only their last four characters on screen, in registers, in exports and in Vericia AI.

Four-eyes approval

KYC decisions, screening sign-offs and changes to rules or fees can each require approval by a second person.

Audit trail

Every change is recorded with its values before and after, who made it and when, together with every occasion a masked number was viewed in full.

In how we build it

Security built into every release

Every change follows a documented secure development policy, with automated security checks on every build. A build that fails a check is not released.

  • Recognised standards. The NIST Secure Software Development Framework, OWASP ASVS 5.0 Level 2, the OWASP Top 10:2025 and NIST SP 800-63B-4 for authentication.
  • Static analysis. Security warnings in the API or the web application stop the build.
  • A controlled supply chain. Every third-party component is pinned to an exact version and checked for known vulnerabilities on every build and every day.
  • Secrets scanning. The full code history is scanned for credentials, and the build pipelines themselves are audited.
  • Automated security testing. Data segregation, rate limiting, security headers and the audit trail are tested with every release.
  • Committed remediation times. Critical vulnerabilities are fixed within 3 days, high within 7 and moderate within 30. Anything listed in CISA's Known Exploited Vulnerabilities catalogue is treated as critical.

Where it runs

Hosted on Microsoft Azure

Encryption in transit

All connections use TLS 1.2 or later, enforced with HSTS.

Encryption at rest

Databases are encrypted with AES-256 using Azure SQL transparent data encryption.

No stored credentials

Services authenticate to one another and to Azure Key Vault with managed identities, not stored passwords.

Backup and recovery

Automatic database backups with point-in-time restore.

Browser protections

A strict content security policy and security headers on every response.

Responsible disclosure

If you believe you have found a vulnerability, please report it in confidence to contact@vericia.net. We will respond promptly.

AI and integrations

AI held to the same controls as everything else

  • Off by default. Enabled only at your firm's request, and only for the roles you choose.
  • Permission-aware. It reads records with the permissions of the person asking, and nothing more.
  • Human approval. It cannot change a record until the person who asked has approved the change.
  • Masked and logged. Sensitive numbers are masked before anything leaves Vericia, and every question is recorded in the audit trail.
  • Connected apps held to the same rules. Claude, ChatGPT and Microsoft Copilot read only as the person using them, and their proposed changes wait for that person's approval.
  • API keys approved twice. Every key belongs to a service account and is approved by Vericia before it can read anything.
  • Never used for training. Anthropic does not train its models on the data Vericia sends. More on AI safeguards
Audit log showing each change with the values before and after

See it for yourself

Your entities. Your jurisdictions. One source of truth.

Book a 45-minute demonstration. Tell us which jurisdictions you administer, and we will show you Vericia at work on structures like your own.