01
Enrich
Add and merge useful Schema.org data, grounded summaries, FAQs, and cleaner bot-facing HTML across your site.
WebPage · Article · FAQPage · Product · Offer · Breadcrumb
Make your website clearer to AI crawlers without rebuilding it.
Already have access? Open Dashboard →recognized crawler request
Detect
Known crawler
GPTBot, ClaudeBot, PerplexityBot, and other recognized agents
Enrich
Grounded content
Schema.org, summaries, FAQs, and cleaner bot-facing HTML
Serve
Approved output
Published from the edge only after configured trust checks
✓Original site for people
✓No origin-code changes
✓Preview and rollback
One edge layer, three outcomes
Make AI readability operational
Move beyond one-time markup projects with a workflow your content and engineering teams can actually manage.
02
Control
Preview generated content, route sensitive claims to review, exclude URLs, regenerate pages, and roll back configuration changes.
Preview · Review · Edit · Approve · Exclude · Roll back
03
Measure
See recognized bot visits, enrichment coverage, top crawled pages, and identifiable traffic referred by AI products.
Bot visits · Coverage · Top pages · AI referrals
Real product output
Useful structure, grounded in the page you already publish
SigilEdge fetches meaningful origin HTML, generates enrichment from that source, and adds only approved output to the bot-facing response.
- ✓Schema.org types are selected for the page, including product and offer data where applicable.
- ✓Summaries and FAQs carry source spans, confidence, and review state.
- ✓Existing JSON-LD can be merged instead of blindly replaced.
Product page enrichmentapproved
Origin signals
Trail Runner Pro
Price, availability, rating, description
Added at the edge
Product + Offer
JSON-LD plus grounded FAQ and summary
{
"@type": "Product",
"name": "Trail Runner Pro",
"offers": {
"@type": "Offer",
"availability": "InStock"
}
}Review state
Approved
Evidence
Linked source spans
Confidence
Above configured threshold
Illustrative specimen based on the shipped enrichment types, not customer data.
The rest of the product
Onboarding, preview, and measurement live in the same dashboard
SigilEdge is deployed through domain onboarding and a DNS change, without modifying the origin application. What happens afterwards is reported from edge requests and identifiable referrers.
Domain onboardingpreview
- ✓Domain addedexample.com
- ✓Ownership verifiedDNS TXT record
- ✓Origin setNo origin code changes required
- ●Preview openSigned preview link on the dev hostname
- ○DNS cutoverPoint DNS when the preview looks right
Before you move DNS
Open the enriched bot-facing view through a signed preview link. Human visitors on the live site continue to receive the original content.
Illustrative specimen based on the shipped onboarding steps, not customer data.
Coverage & AI crawl activitylast 7 days
Recognized crawler requests
4,812
Last 7 days
Pages with approved enrichment
128 of 149
Discovered pages
Identifiable AI referrals
312
Visits from AI products
Requests by recognized crawler group
OpenAI
Anthropic
Perplexity
Other recognized agents
AI referral sources
ChatGPTPerplexityClaudeGemini
Illustrative specimen with sample values for layout, not customer data or benchmark results. Referral sources are identified from inbound referrers, which is separate from crawler detection.
How it works
A controlled path from origin to edge
The first request for an uncached page can queue generation. Approved enrichment is then cached for subsequent recognized crawler requests.
01
Add and verify a domain
Add your domain in the dashboard and prove ownership before anything is generated or served.
02
Preview on the dev hostname
Set your origin and inspect the bot-facing output through the dev hostname while live traffic is untouched.
03
Point DNS when ready
Complete the guided DNS change on your own schedule. No origin code changes, and people keep receiving your original site.
04
Generate from origin HTML
SigilEdge discovers pages and generates grounded structured data, summaries, and FAQs in the background from meaningful source HTML.
05
Approve, cache, and serve
Confidence, entailment, and claim-risk rules decide what publishes automatically. Approved enrichment is cached and served to recognized AI crawlers.
06
Monitor and refresh
Track coverage, crawl activity, and AI referrals in the dashboard, and regenerate pages when your content changes.
Trust is a product surface
Automation with a review lane
SigilEdge is designed to publish low-risk, well-supported enrichment and escalate the rest, not to invent facts behind your team’s back.
Explore Trust & ControlGrounded generation
Generated summaries and answers retain source evidence from the origin page.
Risk-aware review
Volatile claims such as price, stock, policy, and shipping are held for human review.
Configurable thresholds
FAQ and summary confidence thresholds can be managed independently.
Safe operations
Preview links, URL exclusions, regeneration, configuration history, and rollback keep teams in control.
Graceful degradation
If enrichment is absent or generation fails, SigilEdge falls back to origin content, so the request still returns the original site instead of fabricated output.
Catalog readiness
Coverage by page type
Product schemacovered
FAQ enrichmentreviewed
Grounded summariespublished
Illustrative interface specimen; values are not customer performance metrics.
Built for content at scale
A practical fit for e-commerce catalogs
Product teams get richer machine-readable pages without turning every schema or content update into an origin release.
Product detail pages
Add or merge Product, Offer, AggregateRating, and Breadcrumb data from page evidence, including GTIN and MPN identifiers when the origin page exposes them.
Large catalogs
Discover pages from sitemaps, process asynchronously, and regenerate individual URLs or a full domain.
Fast-changing facts
Hold volatile price, stock, shipping, and policy claims for review before publication.
Origin requirement: Product pages must return meaningful HTML or metadata from the origin. SigilEdge is not a browser-rendering or SPA prerendering service.
Measure what is observable
Crawl activity and AI-referred traffic, without pretending they are citations
SigilEdge reports signals available from edge requests and identifiable referrers. It does not claim to see every answer, mention, or citation inside third-party AI products.
Crawl activity
Total bot visits
Recognized AI crawler requests across your domain over the selected period.
Enrichment rate
The share of crawler requests served with approved enrichment.
Visits by AI engine
Crawl activity broken down by recognized crawler and engine group.
Top crawled pages
URLs receiving the most recognized AI crawler activity.
AI referrals and coverage
Total AI referrals
Visits whose referrer can be identified as an AI product.
Referral sources
Which AI products those identifiable referrals arrived from.
Top referred pages
The landing pages AI referrals arrive on most often.
Site coverage and freshness
How much of your discovered site carries approved enrichment, and what is pending review, excluded, or due for regeneration.
Referral sources
Referrals are grouped by the AI product that sent them, alongside the pages they landed on.
- ChatGPT
- Perplexity
- Claude
- Gemini
- Copilot
Illustrative source labels. Referral products vary by traffic, and no sample volumes are shown here.
Coverage and freshness states
Every discovered URL carries a state, so you can see what is live, what is waiting on a human, and what needs regenerating after a content change.
- Discovered
- Approved
- Pending review
- Excluded
- Needs regeneration
Illustrative state labels from the coverage view, not customer data.
What AI referrals are not: AI referrals represent identifiable visits from AI platforms. They are not a complete measure of citation or share of voice, because many AI answers are read without a click and many products do not pass an identifiable referrer.
Early access
See what AI crawlers receive from your site
Apply for an assisted pilot. We’ll review your origin HTML, content scale, and goals before recommending a rollout.
Domain onboardingGuided reviewNo public pricing promise