HomeProductsEDGE INJECTOR™
Product 06 · Edge Deployment

EDGE INJECTOR™
Deploy Without Rebuilding.

EDGE INJECTOR™ deploys your complete AI signal layer, entity graph, and machine-readable files to the Cloudflare network edge — globally, instantly, and without any modification to your existing website, CMS, hosting, or code.

<10msGlobal Latency
300+Edge Nodes
1CNAME Record
24hLive Deployment
The Problem

Signal Files Are Worthless Until They Are Deployed.

The most common reason businesses do not deploy AI signals is not lack of knowledge — it is lack of a safe deployment path. Modifying an existing website carries risk. Changing CMS configuration can break things. Adding schema markup incorrectly creates structured data errors. Updating robots.txt carelessly can block crawlers that should be permitted.

EDGE INJECTOR™ removes all of these risks. It does not modify the existing website. It deploys a parallel AI infrastructure layer at the network edge — between the internet and the origin server — that serves AI signals independently of whatever the website is doing.

The existing website continues to operate exactly as before. The AI signal layer operates alongside it — transparently, permanently, and without creating any dependency between the two systems.

The Architecture

Two Independent Systems. One Domain.

The website and the AI signal layer are completely decoupled. One change to either does not affect the other.

AI Crawler or Browser

Any request to the domain

Cloudflare Edge — EDGE INJECTOR™

Intercepts every request · serves AI signals from KV · injects schema into page responses

↓ (human browser requests only)

Origin Server — Your Existing Website

WordPress · Webflow · Squarespace · custom HTML · any platform · completely unchanged

AI crawlers receive signal files directly from the edge — the origin server is never reached. Human browsers receive page HTML with schema injected at the edge before delivery.

Tenant-First Architecture

Complete Tenant Isolation at Scale

Every business has a tenant-specific key prefix. All signal files stored and served under that prefix. Zero cross-tenant data leakage.

GET /v1/your-domain.com/ai.json → served from KV in <10ms
GET /v1/your-domain.com/entities.json → served from KV in <10ms
GET /v1/your-domain.com/intents.json → served from KV in <10ms
GET /v1/your-domain.com/ai-actions.json → served from KV in <10ms
GET /v1/your-domain.com/allow-lane-matrix.json → served from KV in <10ms

Tenant-first key structure ensures that every business's signals are completely isolated. Onboarding a new tenant is two steps: upload signal files to KV under the tenant prefix, add CNAME record. Nothing else.

What Gets Deployed

Every Signal File. Every Page. Globally.

All files generated by AI SIGNALS™ and KNOWLEDGE GRAPH™ are deployed to the edge in one operation.

llms.txtLLM instruction file — canonical entity declaration and permission layer for large language modelsDomain root
ai.jsonPrimary AI identity file — entity definition, service catalogue, canonical propertiesDomain root
robots.txtUpdated to explicitly permit GPTBot, anthropic-ai, Google-Extended, PerplexityBot, and all major AI crawlersDomain root
entities.jsonComplete entity graph — business, services, team, locations, certifications with relationship mapsDomain root
intents.jsonBuyer intent registry — all service-to-query mappings, negative intents, and intent filtersDomain root
ai-actions.jsonAI action permissions — what AI systems are permitted to do when referencing this businessEdge KV
ai-routes.jsonQuery routing — connects buyer intent to the most relevant page, signal file, or contact routeEdge KV
allow-lane-matrix.jsonPermission layer — which AI systems have access to which signal tiersEdge KV
Schema JSON-LDAll Schema.org markup injected into every page response at the edge — invisible to human visitors, readable by AI crawlersAll pages
Requirements

What You Need. What You Do Not.

What You Need

  • A Cloudflare account — free tier is sufficient
  • One CNAME DNS record at your domain registrar
  • Signal files generated by AI SIGNALS™
  • 30 seconds to add the CNAME record

What You Do Not Need

  • WordPress admin or any CMS access
  • cPanel, FTP, or SSH credentials
  • Database or backend system access
  • Any developer involvement from your team
  • Changes to existing HTML, CSS, or JavaScript
  • Changes to hosting configuration
  • Any plugin installation
Onboarding

Live in Five Steps. Under 24 Hours Total.

01

Cloudflare Account

Create or connect an existing Cloudflare account. Free tier is sufficient for all standard deployments.

5 min
02

CNAME Record

Add one CNAME DNS record at your domain registrar pointing traffic through Cloudflare. Standard configuration — 30 seconds to complete.

30 sec
03

Signal Files Uploaded

All signal files from AI SIGNALS™ and KNOWLEDGE GRAPH™ uploaded to Cloudflare Workers KV under your tenant key. Automated process.

Automated
04

Edge Worker Deployed

Cloudflare Worker deployed — configured to serve signal files, inject schema markup, and enforce AI crawler permissions. Automated.

Automated
05

Verification and Live

All signal files checked for accessibility. All AI crawlers confirmed permitted. Schema injection verified on all pages. Deployment report delivered.

<24h total
Output

Four Deployment Confirmations

AI Layer

Complete Signal Layer Live

All signal files deployed and accessible at domain root and edge endpoint. Every file returns HTTP 200 with correct content-type headers for every major AI crawler.

Edge Delivery

Sub-10ms Global Response

Signal files served from Cloudflare's 300+ edge nodes. Response time under 10ms from every major global location. No origin server dependency.

AI Discovery Assets

All Files Accessible to AI Crawlers

GPTBot, anthropic-ai, Google-Extended, PerplexityBot, and all major AI crawlers confirmed permitted and able to retrieve all signal files.

Infrastructure Report

Full Deployment Confirmation

Structured report confirming every deployed file, every accessibility check result, every crawler permission status, and every schema injection verification.

The Result

Before and After EDGE INJECTOR™

Before
  • Signal files exist locally but not deployed
  • AI crawlers cannot access any signal file
  • Schema markup absent from all pages
  • robots.txt blocks major AI crawlers
  • Origin server must serve every request
  • Website changes required to add any signal
  • No way to deploy without developer involvement
After
  • All signal files live at domain root and edge endpoint
  • Every major AI crawler can access all signal files
  • Schema markup injected into every page at the edge
  • All AI crawlers explicitly permitted in robots.txt
  • Edge serves AI signals in under 10ms globally
  • Website completely unchanged — zero risk
  • Deployed without any developer involvement
Next Step

From Deployment to Trust

Product 07

TRUST LAYER™

With signals deployed and accessible, the next layer is trust. TRUST LAYER™ builds the governance, transparency, and citation signals that move a business from Verifiable™ to Trustable™.

Go to TRUST LAYER™ →
Previous

KNOWLEDGE GRAPH™

EDGE INJECTOR™ deploys the files generated by KNOWLEDGE GRAPH™ and AI SIGNALS™. Complete both before deploying to the edge.

← Back to KNOWLEDGE GRAPH™