Private deployment · Model acceptance · Administrator-managed accessReview the control model

Security and governance

Controls that begin before the model ships.

Deployment, access, data boundaries, validation, evidence, release governance, and operations are defined into the delivery scope.

Release evidenceModel acceptance workflow
23Registered models
18Accepted
3In validation
2Restricted
Source approvalPassed
Performance gatePassed
Bias reviewPassed
Security reviewPassed
5Release approvalPending

Control model

Private by default. Auditable by design.

FiboRisk does not claim a control or certification that has not been contractually defined and independently established. The public website describes the delivery model; the executed SOW controls the customer environment.

Provision a dedicated tenant or agreed private deployment with separated environments, controlled connectivity, and responsibilities defined in the SOW.

  • No public self-service production accounts
  • Environment and connectivity boundaries
  • Customer-specific service responsibilities

Keep approved customer data within the agreed deployment and retention boundary, with access, transfer, encryption, and deletion requirements defined before delivery.

  • Approved sources and processing purpose
  • Encryption and retention requirements
  • Controlled export and deletion paths

Customer administrators invite named authorized employees and manage roles aligned to development, validation, approval, operation, and audit responsibilities.

  • Enterprise identity integration
  • Least-privilege roles
  • Separation of duties and administrator control

Retain the source, feature, model, scenario, policy, decision, action, override, reviewer, approval, and applicable version as one inspectable chain.

  • Versioned evidence objects
  • Exportable acceptance records
  • Change, incident, and release history
01

Model documentation

Purpose, scope, assumptions, limitations, data, features, methods, intended use, and excluded use.

02

Validation evidence

Performance, stability, sensitivity, drift, bias, explainability, limitations, and challenger results.

03

Control evidence

Access, environments, approvals, changes, monitoring, overrides, incidents, and operational ownership.