Current security posture, as of June 4, 2026.
What we do, how we do it, and what your security review team gets in writing. Refreshed on every deploy.
Engineering principles
How we build secure systems.
Not a list of certifications we hold — a list of practices our engineers follow on every engagement.
Defense in depth
Network, identity, application, and data planes each enforce their own controls. No single layer is load-bearing for the system's safety.
Least privilege by default
Every service identity, every user role, and every CI job receives the narrowest set of permissions it needs to do its job. Broad roles require a written justification.
Threat modeling per engagement
Every engagement opens with a written threat model — what we are protecting, who the adversaries are, and how the controls map to them. Updated on material change.
Secrets management discipline
Secrets live in cloud KMS or a managed vault, never in source. CI rotates short-lived credentials, and access is logged at the secret level.
Continuous dependency scanning
Every repo runs dependency and container scanning in CI on every PR. Critical vulnerabilities block merge until remediated or risk-accepted in writing.
Security review gate before production
Every production deploy passes a written security review — control coverage, threat-model deltas, and a named owner — before traffic moves.
Data handling
How we handle your data.
Encryption
- AES-256 at rest across managed cloud stores and customer-controlled BYOK where required.
- TLS 1.2 or higher in transit, with mTLS between service-to-service hops inside the cluster.
- Per-tenant key separation where the data model warrants it; documented per engagement.
Access controls
- Federated SSO with mandatory MFA for all human access to production planes.
- Role-based access control with periodic review and just-in-time elevation for production changes.
- Audit logs retained per regulatory minimum, shipped to a dedicated logging account.
Data residency options
- US-only deployments by default; FedRAMP-aligned regions (Azure Government, AWS GovCloud) for federal workloads.
- Customer-controlled BYOK for tenants that require key custody.
- Single-tenant deployments inside the customer's cloud account where the use case requires it.
Vendor security questionnaire
Our standard vendor security questionnaire response.
A pre-completed response covering the questions enterprise security teams ask most often. Send your own questionnaire to security@theprrgroup.com and we will return our completed response within one business day.
Incident response
How we respond when something breaks.
Detect
Centralized log aggregation, anomaly alerts on identity and data-access patterns, and pager rotation across senior engineers.
Triage
A written playbook per workload class drives the first hour — scope, contain, and confirm impact before we move further.
Notify
We commit to a customer notification SLA in writing in the MSA. Default is one business day for confirmed incidents touching customer data.
Runbooks
Every customer engagement ships with incident runbooks specific to its architecture — what to check, who to call, how to roll back.
Security inquiries.
For vulnerability reports, security review requests, and questionnaire submissions: