Policy Continuum

The highest level of abstraction, expressing requirements in terms of enterprise outcomes and stakeholder value. This layer is completely decoupled from technical implementation.

• Language: Natural language or simple declarative statements.
• Example: 'Ensure platinum-tier customers have a flawless video conferencing experience.'
• Key Function: Captures the 'what' and 'why' before any 'how' is considered.

Translates business goals into technology-agnostic network requirements. This layer defines the performance envelope without specifying vendors or protocols.

• Language: Formal policy rules and service-level objectives (SLOs).
• Example: 'Traffic marked as telephony must have < 150ms latency and < 1% packet loss.'
• Key Function: Defines measurable success criteria for the network service.

Introduces domain-specific logic and vendor-neutral configuration models. This layer maps operational requirements to specific network functions like routing, QoS, and security.

• Language: YANG models, OpenConfig, or other structured schemas.
• Example: 'Apply a strict priority queuing profile to the Differentiated Services Code Point (DSCP) value EF (46).'
• Key Function: Creates a logical device model independent of CLI syntax.

The lowest level of abstraction, consisting of vendor-specific, machine-executable code. This is the literal syntax pushed to physical or virtual hardware.

• Language: CLI commands, NETCONF/YANG payloads, or REST API calls.
• Example: set class-of-service schedulers strict-high-priority transmit-rate percent 20
• Key Function: The final atomic instruction set that directly manipulates hardware ASICs and forwarding tables.

The algorithmic engine that decouples layers, allowing a change at the Business Intent Layer to automatically propagate down without manual re-scripting. This mechanism relies on declarative modeling rather than imperative scripting.

• Process: Intent Translation converts a high-level graph into a low-level syntax tree.
• Benefit: Prevents vendor lock-in, as the same operational intent can synthesize configurations for Cisco, Juniper, or Arista hardware.

A continuous verification stream that flows bottom-up from the Device Layer to the Business Layer. Real-time telemetry is aggregated and validated against the original intent.

• Mechanism: Streaming telemetry (gRPC) feeds into an Intent Assurance engine.
• Drift Detection: If a device configuration drifts due to a manual override, the system flags a violation at the Operational Layer.
• Outcome: Enables closed-loop automation by triggering remediation workflows to restore the intended state.

The algorithmic process of converting a declarative business policy into a set of device-specific, low-level network configurations. This is the critical bridge between the business intent and system intent layers of the continuum.

• Ingests abstract policy models (e.g., YANG, TOSCA)
• Outputs vendor-specific CLI, NETCONF, or RESTCONF payloads
• Often uses policy abstraction to decouple logic from syntax
• Validates feasibility against available resource inventory before deployment

A continuous validation loop that uses real-time streaming telemetry to verify that the network's operational state matches the declared intent. This closes the loop at the bottom of the continuum.

• Compares observed state against Service-Level Objectives (SLOs)
• Detects intent drift caused by congestion, failure, or misconfiguration
• Triggers automated remediation workflows when violations occur
• Provides evidence for intent compliance audits and reporting

An algorithmic mechanism that detects and resolves overlapping or contradictory intents at the translation layer. As multiple business units declare policies, conflicts are inevitable.

• Uses priority-based arbitration: high-priority intents preempt lower ones
• Employs negotiation-based logic for resource allocation trade-offs
• Prevents conflicting QoS guarantees or security postures
• Essential for multi-tenant, multi-domain orchestration environments

The automated generation of correct-by-construction, low-level device configurations from a high-level intent model. This represents the final stage of the continuum before device actuation.

• Often leverages formal methods to guarantee syntactic and semantic correctness
• Eliminates manual CLI errors that cause outages
• Synthesizes configurations for heterogeneous, multi-vendor hardware
• Enables zero-touch provisioning at scale

A precise, measurable performance metric defined within an intent that the closed-loop system must continuously maintain. SLOs are the quantifiable bridge between business intent and technical implementation.

• Examples: 99.999% availability, sub-10ms latency, 0% packet loss
• Monitored by the intent assurance function
• Violations trigger automated remediation or alerting
• Distinct from broader SLAs which include business penalties

A formal model representing the lifecycle stages of a network intent and the valid transitions between them. This governs the policy continuum's operational behavior.

• Stages: Creation → Validation → Translation → Fulfillment → Assurance → Decommissioning
• Prevents invalid state transitions (e.g., deploying an unvalidated intent)
• Tracks intent versioning and modification history
• Provides audit trail for governance and compliance