Methodology - How Lumet Scores AI Visibility

How it works

Each tier builds upon the previous, creating a complete AI visibility pipeline from analysis to optimization.

Understand your content landscape and identify gaps in AI visibility.

Step 1

Deep web scraping discovers all your content

Crawl4AI extracts markdown, metadata, and page structure

Step 2

Content is chunked and vectorized

500-token chunks with 50-token overlap, 1536-dim vectors

Step 3

Semantic search against buyer queries

pgvector similarity search identifies content gaps

Step 4

Multi-dimensional gap analysis

Coverage, extractability, citation & geo readiness

Your WebsiteHTML, JSON-LD

Edge FunctionsDeno Runtime

pgvector1536-dim embeddings

Gap ReportActionable insights

How completely your content addresses buyer search queries.

How easily AI systems can parse and extract information.

Likelihood of being cited in AI-generated responses.

Local SEO signals for location-aware AI queries.

Each tier builds on the previous. Generation requires Intelligence data, and Measurement tracks content from both lower tiers.