Our AI SEO Technology Stack

Purpose-built tools that solve specific problems in AI search optimization. Proprietary systems we developed, use daily in client work, and continuously improve.

We built these because they didn't exist. The AI SEO market is young enough that agencies need tools established vendors haven't created yet.

AI Bot Tracking & Visibility Monitoring

BotSight

React Tauri / Rust Edge Middleware MCP Server

Desktop application that monitors AI crawler activity in real time. Deploys lightweight edge middleware on client sites and provides a full analytics dashboard with 30-day trends, scoring, coverage gap analysis, and exportable reports.

15+ AI Bots Tracked
4 x 25 Visibility Score
17 MCP Tools

Crawlers Tracked

  • ChatGPT (GPTBot, ChatGPT-User)
  • SearchGPT (OAI-SearchBot)
  • Claude (ClaudeBot, Claude-User, Claude-SearchBot)
  • Perplexity, Google AI, Meta AI, TikTok AI
  • Amazon Q, Apple AI, DuckDuckGo AI
  • Indirect: Gemini (via Google), Copilot (via Bing)

Key Capabilities

  • AI Visibility Score (Bot Diversity, Crawl Frequency, Recrawl Rate, Page Coverage)
  • GSC, Bing Webmaster, GA4 integration
  • Per-page search correlation analysis
  • Security threat detection
  • PDF reports and Excel workbooks
  • AI agent integration via MCP tools

Used in our AI Visibility Monitoring service.

Industry-Agnostic AI Query Generation

AI Query Predictor

Python GSC / Bing API 97K Training Queries

Transforms real search console keywords into AI-native audit prompts that mirror how consumers actually query chatbots. Auto-derives user context and industry from the keywords themselves — works for any vertical, B2B or B2C, with no manual configuration.

Intent Classification

  • SERVICE — product/service focused (highest value)
  • INFORMATIONAL — guides, comparisons
  • BRANDED — brand-specific queries
  • NAVIGATIONAL — login, docs
  • IRRELEVANT — API calls, code snippets (filtered)

Output Features

  • 3+ prompt variations per keyword with quality scoring
  • Context auto-derived from keywords (no hardcoded industry pools)
  • 5 intent types: Informational, Discovery, Validation, Comparison, Review-Seeking
  • Trained on 97K ESCI queries + 341 verified chatbot queries

Powers our Competitive Intelligence and GEO Content Strategy services.

Cross-Platform AI Citation Extraction

Multi-Platform Scraper

Python Patchright SSE Interception DOM Extraction

Autonomous system that runs queries across four AI platforms with isolated browser sessions per account, eliminating personalization bias and producing statistically meaningful citation data.

40 Sessions / Query
4 AI Platforms
3-Tier Auth System
  • ChatGPT, Claude, Perplexity, and Google AI Mode -- 10 sessions each per query
  • SSE stream interception (Claude, Perplexity) and DOM extraction (ChatGPT, Google AI Mode)
  • Session restore (~5s), re-auth (~30-60s), or full account creation (~2-4 min) fallback
  • Per-account daily limits, rate limiting, and automatic cooldowns
  • Technical GEO audit of top-cited URLs (schema, headings, readability, persuasion patterns)
  • Fan-out query extraction -- the web searches AI platforms trigger during response generation

The engine behind our Competitive Intelligence service. All analysis grounded in the GEO Knowledge Base.

Empirically Validated Optimization Intelligence

GEO Knowledge Base

KuzuDB PostgreSQL e5-small-v2 MCP Server

Dual-layer retrieval system combining vector embeddings with a knowledge graph. Continuously updated as AI platform behaviors change, and exposed as an MCP server for direct AI agent access during audits.

341 Optimization Chunks
3,555 Graph Triplets
6 Domains
4 Search Modes

Domains Covered

  • ChatGPT Optimization
  • Claude Optimization
  • Perplexity Optimization
  • Google AI Mode Optimization
  • SEO Best Practices
  • Technical GEO Standards

Search Modes

  • Hybrid (70% vector + 30% full-text) -- recommended default
  • Pure vector (semantic similarity)
  • Full-text (PostgreSQL tsvector, exact terms)
  • Graph (entity relationship traversal)

Every recommendation in our AI SEO Audits and GEO Content Strategy deliverables is cross-referenced against this knowledge base.

Free and Open Source

Open Source Tools

Orunla

Rust CLI MCP Server

Persistent memory system for AI agents. Gives AI coding assistants a knowledge graph that remembers facts, relationships, and context across sessions.

CGC

Python MCP Server

Context Graph Connector -- maps and chunks files, databases, and spreadsheets for efficient context window usage in AI workflows.

Learn more about our open-source projects →

Limitations & Context

Complementary, not replacement. These tools cover AI-specific gaps that platforms like SEMrush, Ahrefs, and Google Search Console don't address. They are not general-purpose SEO tools.
Solo researcher/developer. Rapid iteration and deep domain expertise, but development pace is constrained. We prioritize features that directly impact client outcomes.
Current as of March 2026. AI search is a fast-moving field. Our tools and knowledge base reflect the current state of AI platform behavior, and we update methodology as platforms evolve.

Want These Tools Working for Your Agency's Clients?

Our tools power our AI SEO services. Tell us about your agency and we'll show you how they can help your clients.

Talk to Us
Published research → From the blog → About the founder →