Information Security Policy
The written policy governing how this platform protects customer data and its own systems: who is responsible, what is required of everyone with access, and what happens when something goes wrong.
1. Scope and ownership
This policy covers every system that stores, processes, or transmits customer data on this platform: the control-plane API, the portal and admin applications, the background bot workers, the database, and the infrastructure they run on.
| Role | Responsibility |
|---|---|
| Security owner | Engineering Lead, Orenyx Engine development team — self-assigned as interim owner. Accountable for this policy, approves exceptions, is the point of contact for a reported vulnerability or breach, until the client names a permanent owner on their side. |
| Engineering | Implements and maintains the technical controls in section 3, and reports a suspected incident to the security owner immediately, not after investigating it themselves first. |
| Every team member with system access | Follows the access-control and acceptable-use rules in sections 2 and 4, and is the first line of defense: reporting anything that looks wrong costs nothing, staying quiet about a real problem does not. |
2. Access control
- Least privilege by default: a new team member is given the lowest role tier (Analyst/Auditor or Developer) that lets them do their job, not Admin or Bot Owner by default. See the Admin Console guide for what each tier can do.
- Two-factor authentication is available to every account from Settings and is strongly recommended for anyone with Admin, Bot Owner, or Developer access.
- API keys are scoped to only the parts of the system a given integration needs, and expire automatically after 90 days.
- Access is reviewed whenever someone leaves a role or a project ends, at minimum. A recurring quarterly review of who has access to what is the target cadence; this is a commitment to add to the team's calendar, not yet an automated or tracked process.
- Departing team members have their access revoked the same day their role ends, this is a manual step today (removing their user record and revoking their API keys/sessions) and should be added to an offboarding checklist.
3. Data protection controls already in place
These are real, verified controls, not aspirational, each one is described in full detail elsewhere in this documentation.
| Control | Where it is described |
|---|---|
| Encryption of sensitive lead fields at rest (AES-256-GCM) | SOC 2 Overview |
| HTTPS enforced for all traffic | SOC 2 Overview |
| Mandatory or optional two-factor authentication per account | Authentication |
| Tenant data isolation, verified against cross-account access attempts | SOC 2 Overview |
| Tamper-evident financial ledger with hourly automatic verification | Settlement System |
| Append-only audit logging of sensitive actions | Admin Console |
| Daily automated database backups | SOC 2 Overview |
4. Data classification
| Category | Examples | Handling requirement |
|---|---|---|
| Restricted | Passwords, API key secrets, Stripe payment identifiers, encryption keys | Never logged, never emailed in plaintext, never stored outside the fields specifically designed to hold them. |
| Confidential | Lead contact details and messages, account holder information, ledger and settlement records | Access limited by role and tenant isolation. Exported or downloaded copies (CSV/JSON) should be deleted once no longer needed. |
| Internal | Audit logs, usage metrics, non-customer configuration | Visible to authenticated team members appropriate to their role, not published externally. |
| Public | This documentation, marketing pages, the status page | No restriction, written for external readers by design. |
5. Incident response
When a security incident is suspected, in rough order:
- Contain it first: revoke the specific API key, session, or credential involved, or suspend the affected account, before doing a full investigation. A live, ongoing compromise being investigated slowly is worse than a fast, imperfect containment.
- Notify the security owner immediately.
- Preserve evidence: the audit log and ledger are append-only by design specifically so an incident can be reconstructed after the fact, do not delete or alter records while investigating.
- Assess whether customer data was affected. If it was, follow the notification approach described in the Data Processing Agreement.
- Write a postmortem once resolved, in the same honest, plain-language format used for the incidents published on the Status & Support page, what happened, what it affected, and what changed to prevent it recurring.
6. Third-party vendors
Every vendor that touches customer data is listed, with what they process, on the Compliance page. Adding a new vendor that will touch customer data should update that list as part of the same change, not as an afterthought.
7. Review
This policy is reviewed at least quarterly by the engineering team, or immediately after any security incident, whichever comes first. That cadence is self-imposed by the team maintaining the platform, not yet a client-mandated or externally audited process, that distinction matters and is not being hidden.
Self-review log: the development team has conducted its own internal review, covering tenant data isolation testing, dependency and static-analysis checks, authentication and access-control paths, and the encryption/backup controls described in section 3. This is a self-assessment by the people who built the system, useful for catching real gaps early, but it is not a substitute for an independent third-party audit. That remains the outstanding step described in the SOC 2 Overview.
Was this page helpful?