Structured knowledge base + programmatic SEO

The problem

Roofing information online is fragmented and largely unsourced. Materials, products, manufacturers, suppliers, and permitting rules live in scattered pages with no way to tell where a claim came from. There is no clean, provenance-tracked reference to build on.

What it does

A structured database of roofing materials, products, manufacturers, suppliers, and permitting authorities, held to strict provenance. An 8-pass validation pipeline checks and sources the data before it is emitted as Schema.org JSON-LD across 230 interlinked, search-optimized pages. The content compounds like an asset instead of aging like a brochure.

• Provenance-disciplined data: every fact traces back to a source
• 8-pass validation pipeline before anything ships
• Schema.org JSON-LD structured-data output
• 230 programmatically generated, interlinked pages, live in Google's index

How it was built / my role

Architect. TypeScript throughout, with SQLite and FTS5 full-text search as the data layer and JSON-LD as the structured-data output. Pages are generated programmatically from validated records, so the SEO surface grows with the data instead of with hand-written HTML.