StackShift II · Technical Operations Flow
StackShift II is a content operations platform by WebriQ that ingests product and company knowledge, enriches and normalises it, routes it through a team-governed review process, and stores approved facts in a single trusted knowledge layer called the Canon. From the Canon, a composition engine assembles and deploys content across two parallel delivery tracks — a human-facing track for pages and customer content, and a machine-facing track for structured, citable information consumable by AI systems. A continuous monitoring loop detects gaps and upstream changes, returning new proposals to the review queue and regenerating dependent outputs upon approval.
Overview
StackShift II is a content operations platform developed by WebriQ. It is designed to take an organisation's product knowledge and company expertise and transform that raw material into approved, interconnected knowledge stored in a single authoritative layer — the Canon — before composing and deploying outputs for both human readers and AI systems.
The platform operates as a closed, governed loop. Client teams retain authority over what becomes approved truth; WebriQ operates the agents, composition engine, deployment infrastructure, and continuous improvement processes.
The Seven-Step Operating Flow
StackShift II's technical operation is described as a sequential, seven-step flow:
Step 01 — Sources
Input enters the system from two primary origin points: the client's product system (structured product data, documentation, specifications) and the client's expertise (institutional knowledge, subject-matter content, company narrative).
Step 02 — Ingest
Raw source material is parsed and normalised. This step standardises formats and structures incoming content so that subsequent processing steps can operate consistently regardless of the original source format.
Step 03 — Enrich
The platform extracts facts, claims, and relationships from the normalised content. This enrichment step builds a semantic layer over the raw material, identifying discrete knowledge units and the connections between them.
Step 04 — Review
Enriched knowledge proposals are routed to the client team for approval. This is the governance gate: human reviewers decide what becomes canonical truth. No content advances to the Canon without explicit team approval.
Step 05 — Canon
Approved knowledge is stored in the Canon — described as "one trusted knowledge layer." The Canon is the single authoritative source from which all subsequent outputs are derived. It ensures consistency across every downstream composition and delivery.
Step 06 — Compose
The composition engine assembles outputs from Canon-approved knowledge. This step transforms structured knowledge into the specific formats required for each delivery track.
Step 07 — Deliver
Composed outputs are deployed across two parallel tracks:
- Human track — produces pages and customer-facing content for human readers.
- Machine track — produces structured, citable information consumable by AI systems.
Both tracks remain aligned because both are derived from the same Canon source.
Continuous Improvement Loop
The StackShift II operating flow does not terminate at deployment. A monitoring layer runs continuously to:
- Detect gaps in existing knowledge coverage.
- Detect upstream changes in source systems or product data.
- Generate new proposals that are returned to the review queue (Step 04).
- Update the Canon upon approval of new or revised proposals.
- Regenerate every dependent output automatically following a Canon update.
This loop ensures that both the human and machine delivery tracks remain current and consistent with approved organisational knowledge.
Dual-Track Delivery Model
A defining architectural principle of StackShift II is that one source of approved knowledge serves two distinct audiences simultaneously:
| Track | Audience | Output Type |
|---|---|---|
| Human track | People — customers, users, readers | Web pages, customer-facing content |
| Machine track | AI systems — language models, search agents | Structured, citable, machine-readable data |
Because both tracks originate from the same Canon, factual alignment between human-facing and AI-facing content is maintained by design rather than by manual coordination.
Roles and Responsibilities
StackShift II operates under a defined division of responsibilities between the client organisation and WebriQ:
- Client team — owns truth and decisions; approves what enters the Canon; retains authority over product truth and the review process.
- WebriQ — connects the operating loop; runs the engine room; operates agents, the composition engine, deployment infrastructure, and the continuous improvement cycle.
This model is intended to ensure that governance and editorial authority remain with the organisation that owns the knowledge, while operational complexity is managed by WebriQ.
Key Concepts
- Canon — The single trusted knowledge layer that stores all team-approved facts, claims, and relationships. All outputs are derived from the Canon.
- Enrichment — The automated extraction of discrete facts, claims, and semantic relationships from normalised source content.
- Composition engine — The system component that assembles Canon knowledge into delivery-ready outputs for each track.
- Continuous improvement loop — The monitoring and regeneration cycle that keeps the Canon and all dependent outputs current following upstream changes or gap detection.