Docs · Architecture

Architecture overview

Platform Intelligence v2 turns the live site into a maintainable digital ecosystem with one shared runtime for health analysis, reporting, discovery, and recommendations.

Principles

  • Extend, do not redesign: homepage identity and completed work remain intact.
  • One ecosystem: every product reuses shell, tokens, cards, and SEO patterns.
  • No duplicate sources of truth: registry metadata and knowledge-graph relationships own cross-cutting concerns.
  • Static-first: GitHub Pages compatible, minimal JavaScript, progressive enhancement.
  • Safe by default: Device and Downloads content stays lawful guidance and transparent assets only.
  • Shared engines first: health capabilities must plug into reusable runtime systems before introducing new logic.

Product architecture

Core site

Home, About, Services, Research, Projects, Blog, Credentials, Contact.

Products

AI, Device Service, Downloads, Labs, Docs — plus Research, Open Source, Capabilities, Projects.

Future slots

Academy, Community, Console, API, Cloud — reserved, not shipped.

Folder structure

/assets
  /css /js /images /icons /data /downloads /components
/ai /device /downloads /research /projects /docs /labs
/open-source /capabilities
/templates /schema
/scripts

Canonical Device Service path remains /device/ to preserve live links. Product name in UI is “Device Service”.

Design system

Tokens and components live in assets/css/main.css. Product-specific polish (AI tools) may add a thin layer such as ai-platform.css without forking the brand.

  • Glass cards, premium typography, consistent spacing
  • Accessible contrast on the dark theme
  • Minimal motion; reduced-noise interactions
  • Future light mode can flip CSS custom properties without markup rewrites

Data layer

  • /assets/data/products.json — product registry and nav roles
  • /assets/data/search-index.json — global client-side search corpus
  • /assets/data/downloads.json — download catalog + safety policy
  • /assets/data/capabilities.json — capability explorer catalog
  • /assets/data/platform-registry.json — single metadata source of truth for Platform Intelligence capabilities
  • /assets/data/knowledge-graph.json — relationship layer between registry items

Shared runtime

Platform Intelligence v2 standardizes five reusable engines: Health, Recommendation, Knowledge, Search, and Report. Website Health is the first live capability on this shared flow, and future Device Health, Project Health, and AI Workflow Health capabilities must reuse the same contract.

Canonical flow: data collection → Health Engine → Recommendation Engine → Report Engine → shared report.

Migration rules: legacy URLs stay valid through redirects, canonical health paths live under /intelligence/, and duplicated report rendering is removed as pillars migrate.

SEO & accessibility

Every public page targets title, description, canonical, Open Graph, Twitter Card, JSON-LD (including breadcrumbs), semantic landmarks, keyboard access, visible focus, and WCAG AA contrast.

Performance

No SPA framework. Shared CSS/JS are cacheable. Images use explicit dimensions where present. Search and shell mount progressively. Target: Lighthouse 95+ on content pages with lean assets.

Scalability

New capabilities become canonical paths under /intelligence/ plus registry entries, knowledge-graph relationships, and shared engine integration. Search grows through the JSON index instead of hard-coding every page script.

Roadmap after this foundation

  • v4.0 — AI Platform depth (agents reserved until catalog/docs mature)
  • v4.5 — Device Service interactive decision trees
  • v5.0 — Downloads hub filters and printable assets
  • v6.0 — Workspace dashboard concept
  • v7.0 — Optional cloud/auth only if leaving pure static hosting

Back to Docs Component library