Introduction

What is Lociator?

Lociator derives from locus (Latin: "place", "point", "space") and the suffix -ator (one who operates, one who executes). Lociator = The one who operates loci — the orchestrator of knowledge points in space.

In the age of AI and GEO (Generative Engine Optimization), content is no longer a collection of isolated pages. Every article is a knowledge node. Every topic is a semantic cluster. Every website is a topology map. Lociator was born to orchestrate that map.

It is a web-based SEO analysis platform that crawls your website and builds an interactive graph of its internal linking structure. It combines BFS crawling with AI-powered topic analysis to help you understand page relationships, discover orphan content, optimize link equity distribution, and identify topical clusters across your site.

Key Benefits

Interactive graph visualization: Explore your site's structure through four layout modes — Rings, Tree, Force, and Silos — each revealing different structural insights.
Orphan page detection: Instantly find pages with zero incoming internal links that are invisible to search engine crawlers.
Depth analysis: Measure click distance from the homepage to every page using accurate BFS-based depth calculation.
6-factor architecture scoring: Get a weighted composite score covering depth, linking density, silo structure, orphan ratio, pillar page presence, and cross-silo connectivity.
AI topic clustering: Automatically detect topical clusters and hierarchical content relationships using LLM embeddings.
Smart content extraction: Extract meaningful page content using semantic noise removal — keeping headings and body text while stripping navigation, sidebars, and widgets.
Configurable exclude patterns: Skip specific URL patterns during crawling via user settings.

Who is it for?

SEO professionals who need to audit internal link structures at scale and identify architectural weaknesses.
Content strategists who want to verify topical clusters are properly linked and discover content gaps.
Web developers who need to validate navigation paths, sitemaps, and page discoverability.
Site owners who want a clear, visual picture of how their website is structured and where to improve.

How It Works

The platform uses a Breadth-First Search (BFS) algorithm with concurrent fetching (5 parallel requests, 200ms inter-batch delay) to systematically crawl your website from the root URL. Each page is parsed with Cheerio to extract internal links, page titles, and content text. URLs are normalized (removing tracking params, fragments, trailing slashes) and filtered (skipping nofollow links and 40+ asset file types).

After crawling completes, the system builds a directed graph in the database (nodes = pages, edges = links), computes in-degree/out-degree metrics, identifies orphan pages, and calculates a comprehensive architecture score. Finally, an AI topic analysis pipeline generates topical clusters from the extracted content using vector embeddings.

System Architecture

Lociator uses a distributed architecture:

Frontend: Next.js static site deployed on Cloudflare Pages, with D3.js for graph visualization and Zustand for state management.
API Worker: Cloudflare Worker handling API routes (/api/crawl, /api/status, /api/result, /api/jobs) and topic analysis.
Crawler: NestJS stateless server receiving jobs via Upstash QStash HTTP webhooks, deployable behind Traefik for horizontal scaling.
Database: Supabase (PostgreSQL) for auth, crawl metadata, graph data (pages/links), metrics, topic clusters, and user settings.

Next Steps

Ready to get started? Head over to the Quick Start Guide to run your first crawl, or explore Core Concepts to understand the fundamentals.