Security / Trust Architecture

Security engineered into the operating system.

Quantos is designed for enterprise, institutional, defence, and other high-consequence operating environments. Security is treated as an architectural property: identity-bound access, tenant separation, controlled execution, evidence preservation, and accountable change.

01

Security principles

Quantos is built around the assumption that operational intelligence must remain attributable, reviewable, and contained within the authorised environment. Controls are selected to preserve confidentiality, integrity, availability, traceability, and institutional accountability.

01.1

Least privilege by default

Access is limited to the minimum authority required for an approved role or system function. Elevated access is restricted, accountable, and reviewed according to the deployment model.

01.2

Deterministic control paths

Critical platform operations are designed to follow defined control paths rather than opaque, unbounded execution. System behaviour is intended to remain inspectable and attributable.

01.3

Evidence before assertion

Operational actions, outcomes, scoring, learning, and system corrections are designed to generate evidence that can support review, investigation, and governance.

01.4

Environment-specific enforcement

Security controls are mapped to the deployment boundary, customer requirements, data sensitivity, operational criticality, and contractual obligations.

02

Platform security architecture

Quantos is positioned above existing systems as a closed-loop decision-control layer. Its security architecture is designed to preserve separation between identity, tenant context, data ingestion, execution, evidence, and administrative control.

Trust boundary
Each customer deployment is governed by a defined environment boundary, approved integrations, authorised identities, and explicit data flows.
Tenant context
Platform operations are bound to an authorised tenant or system identity. Tenant context is treated as a required execution condition rather than an optional application parameter.
Execution control
Jobs, workers, scheduled tasks, and internal control functions are designed to execute through explicit identities, approved queues, and auditable orchestration paths.
Evidence plane
Security-relevant and operationally significant events are recorded to support traceability across signal, decision, action, outcome, scoring, learning, and correction.
Deployment profiles
Depending on the engagement, Quantos may be deployed in managed cloud infrastructure, a customer-controlled private environment, or an isolated air-gapped environment.
03

Identity and access control

Human and machine access is governed through identity, role, purpose, and environment. Administrative capability is separated from ordinary use, and access changes are expected to follow controlled provisioning and removal processes.

  • Strong authentication. Administrative access uses strong credential controls and, where supported by the deployment environment, multi-factor authentication.
  • Role-based authority. Access is assigned according to approved responsibilities and restricted to the functions and data required for that role.
  • Machine identity. Internal services, scheduled jobs, and system functions are expected to operate through explicit service or system identities rather than anonymous execution.
  • Lifecycle control. Access provisioning, review, modification, and removal are handled through controlled processes intended to reduce stale, excessive, or orphaned privileges.
  • Privileged access accountability. Elevated operations are restricted and logged according to the capabilities of the contracted deployment.
04

Data protection and isolation

Quantos applies purpose limitation, data minimisation, controlled ingestion, tenant-aware execution, transport security, and environment-appropriate storage protections.

04.1

Purpose-bound processing

Customer data is processed to deliver the contracted service and is not repurposed outside the agreed scope without appropriate written authorisation.

04.2

Tenant separation

Customer environments are designed to preserve tenant separation through deployment, schema, identity, and execution controls appropriate to the contracted architecture.

04.3

Controlled ingestion

Inbound data is expected to pass validation, schema, identity, and integrity checks before it becomes available to downstream control processes.

04.4

Encryption

HTTPS/TLS protects data in transit. Encryption at rest is applied according to the infrastructure, storage architecture, customer requirement, and signed agreement.

04.5

Retention and deletion

Retention is aligned to operational, security, legal, and contractual requirements. Customer-specific retention and deletion terms are defined in the applicable agreement.

04.6

Data locality

Hosting region, deployment location, and data-residency requirements are addressed through the agreed deployment profile and customer security review.

Public disclosure boundary

Specific algorithms, key-management configurations, infrastructure topology, network rules, and environment secrets are not published on this page. They are reviewed under controlled disclosure where appropriate.

05

Secure engineering and operations

Security is maintained through controlled change, dependency management, vulnerability handling, logging, review, and operational discipline across the software lifecycle.

  • Controlled change. Production-impacting changes are expected to pass review, testing, and authorised deployment processes.
  • Dependency risk. Software dependencies and platform components are monitored and updated according to operational risk and compatibility requirements.
  • Vulnerability management. Identified vulnerabilities are assessed, prioritised, and remediated according to severity, exploitability, exposure, and system criticality.
  • Secrets handling. Credentials, tokens, and environment secrets are kept outside public source and are managed through deployment-appropriate secret controls.
  • Logging and monitoring. Security-relevant events and anomalous conditions are logged and reviewed according to the operating context and available telemetry.
  • Separation of duties. Administrative, engineering, deployment, and customer responsibilities are separated where the operating model requires it.
06

Availability, resilience, and recovery

Availability objectives and recovery controls depend on the deployment profile and contracted service level. The platform is designed to support operational continuity through controlled execution, observability, recoverable state, and environment-specific backup practices.

06.1

Operational isolation

Workloads, queues, workers, scheduled jobs, and data stores can be isolated according to customer, environment, and criticality requirements.

06.2

Failure containment

Control paths are designed to prevent a failed task or invalid input from silently becoming an accepted downstream state.

06.3

Backup and restoration

Backup, restoration, recovery-point, and recovery-time requirements are defined by the hosting model and applicable Customer Agreement.

06.4

Air-gapped continuity

Where deployed in isolated environments, Quantos can be operated without dependence on public internet services, subject to the agreed architecture and customer infrastructure.

07

Incident response and investigation

Security incidents are handled through a defined process intended to establish scope, contain exposure, preserve evidence, restore controlled operation, and reduce recurrence.

  • Triage. Reports and alerts are assessed for authenticity, severity, affected environment, and immediate containment needs.
  • Containment. Access, integrations, workloads, or affected components may be restricted to prevent expansion of the incident.
  • Investigation. Available logs, system records, deployment evidence, and customer-provided information are reviewed to establish cause and impact.
  • Recovery. Services are restored through controlled remediation and validation appropriate to the incident and deployment.
  • Notification. Customer notification obligations, timelines, and contact paths are governed by the Customer Agreement and applicable law.
  • Post-incident correction. Confirmed incidents are reviewed to identify required control, engineering, process, or operating changes.
08

Shared responsibility

Security outcomes depend on both Quantos controls and the customer environment. The exact responsibility boundary is documented for each engagement.

  • Customer source systems. Customers remain responsible for the security, integrity, and lawful operation of systems that supply data to Quantos.
  • Identity governance. Customers are responsible for approving users, roles, and access within their organisation and for timely removal of access.
  • Endpoints and networks. Customer-managed endpoints, networks, identity providers, and infrastructure remain under customer control unless explicitly included in scope.
  • Data authority. Customers must have lawful authority to provide data and must obtain any required notices, permissions, or consents.
  • Operational response. Where human action is required, customers remain responsible for timely review, approval, and execution within their operating environment.
09

Responsible disclosure

Security researchers and users should report suspected vulnerabilities privately and responsibly. Do not perform destructive testing, denial-of-service, social engineering, data exfiltration, persistence, or access beyond what is required to demonstrate the issue.

Include
Affected URL or component, reproduction steps, observed behaviour, potential impact, and supporting evidence.
Avoid
Public disclosure before remediation coordination, destructive activity, service disruption, privacy invasion, or accessing unrelated data.
Response
Valid reports are reviewed and acknowledged according to severity, available evidence, and operational context. Remediation timing depends on confirmed risk and affected deployment.
Controlled security review

Review architecture, evidence, and deployment controls.

Prospective and active customers may request a controlled security briefing. Detailed architecture, environment controls, sub-processor information, and available evidence are shared under appropriate confidentiality terms.