intelligentAEO — Intelligence-driven AEO at scale

Analysis Pipeline

Bottom-up analysis,
four levels deep

Every insight is grounded in real page data. Individual page signals compound upward through section synthesis to site-wide strategy — each level informed by the one below it.

Crawl

Playwright: 28+ fields, AI readiness scoring, JS render ratio.

Technical Audit

30+ rules-based checks. AI Readiness 0–100. No API calls.

Keywords

Gemini extraction → clustering → DataForSEO rankings.

Content Briefs

Gemini: keyword, heading outline, rank-aware guidance.

Section Insights

One Gemini call per section. Funnel-stage aware synthesis.

Cross-Section

Site-wide patterns, priority actions, executive summary.

Report

Single-file HTML. Offline-safe. All data in one place.

Agent Library

Every capability
is a registered agent

The agent registry is the backbone of intelligentAEO. Every collector, analyzer, and reporter is a discrete, composable unit — with declared inputs, outputs, cost, and runtime. Activate the right tool for the job, chain them together, and track what's run for every site.

Site Crawler collector

Crawls all site pages using Playwright. Captures 28+ fields including LCP, word count (raw vs rendered), schema types, H1/H2 structure, bot directives, and 11 AI Readiness signals.

url_list crawl_latest.parquet

Keyword Extractor collector

Uses Gemini to extract target keywords from crawl data — title, headings, body text, entity density. Seeds the keyword store with source tagging. Filters by minimum word count and section.

crawl_latest.parquet keywords.json

Keyword Rankings collector

Fetches current SERP positions for all active keywords via DataForSEO. Tracks rank over time. Operates on active keywords in the keyword store — run after seeding and clustering.

keywords.json rankings_latest.json

GSC Collector collector

Fetches Google Search Console performance data — clicks, impressions, CTR, position — per page. Source-tags keywords discovered via GSC. Unlocks Level 4 analytics overlay.

GSC credentials gsc_latest.parquet

Technical Audit analyzer

Rules-based analysis across 30+ on-page checks — title length, meta descriptions, H1 count, schema markup, OG tags, Core Web Vitals, AI bot access, and AI Readiness scoring (0–100, A–D grade).

crawl_latest.parquet technical_audit.parquet

Keyword Cluster analyzer

Groups keywords by topic and intent using Gemini. Processes in batches, reuses existing cluster names for consistency, normalizes naming conventions. Produces cluster summary with intent breakdown.

keywords.json keyword_clusters.json

Gemini Insights analyzer

Four-level bottom-up analysis chain. Level 1: per-page audit with SERP context. Level 2: funnel-aware section synthesis. Level 3: cross-section pattern detection. Level 4: analytics overlay (requires GSC/GA4).

crawl_latest.parquet GEMINI_API_KEY gemini_insights.json

AI Visibility analyzer

Runs configured prompts across Awareness, Consideration, and Decision funnel stages. Scores brand mention, sentiment, citation type, competitors present, and visibility 0–100 per prompt. Pluggable AI source.

ai-visibility-prompts.json ai_visibility.json

HTML Report reporter

Renders the full-site SEO audit as a standalone dark-theme HTML report. Per-URL drill-downs with 4 tabs: Content Strategy, Competitive Analysis, Technical Health, Run History. Self-contained, offline-safe.

All parquet + JSON outputs audit_report.html

Deep Dive pipeline

Full competitive analysis pipeline for a single URL. Discovers top SERP competitors, crawls their pages, runs gap analysis, and produces a standalone deep dive report with actionable strategy.

--url parameter deepdive_report.html

⚡

The agent registry is how intelligentAEO scales. Each agent declares what it needs, what it produces, and what it costs — making it trivial to activate the right combination for a specific audit need, chain agents together in a pipeline, and track output freshness per site across the command center.

Command Center

One dashboard.
Every site.

The command center is where configuration, pipeline status, and AI visibility management come together. Each site gets a full card with live data freshness, analytics integration status, and direct access to every configuration layer.

Site Dashboard

Full-width site cards

Every site gets a card showing configuration status, analytics integration (GA4 and GSC independently), data freshness across all pipeline outputs, and keyword + prompt counts at a glance.

Config status — env, sections, funnel, brand voice

Analytics integration — GA4 and GSC separately

Data freshness — crawl, technical, briefs, insights, cross-section

Keywords seeded · AI prompts configured

Direct links to report and config editor

Agent Registry

Filter, activate, track

The agent registry shows every available agent with its category, description, required inputs, outputs, estimated cost, and runtime. Filter by site to see per-site output status inline.

Filter by category — collector, analyzer, reporter, pipeline

Filter by site — status column appears inline

Requires column — know exactly what to configure first

Cost + runtime estimates per agent

Per-Site Config Editor

Everything in one place

Each site has a tabbed config editor: Brand Voice, Sections, Funnel, Competitors, Keywords, and AI Visibility Prompts. Changes save immediately to config files on disk.

Brand voice — informs all content recommendations

Sections + funnel — maps URLs to stages

Competitors — typed (direct, editorial, listicle) and scoped

Keywords — source-tagged, intent-classified, clustered

AI Visibility Prompts — add, edit, delete per funnel stage

AI Visibility Prompts

Manage and measure

The AI Visibility tab shows the last run summary — score, brand mention rate, per-stage scores — plus the full prompt list grouped by funnel stage. Add new prompts, see results, link directly to the full report.

Last run date, avg score, brand mention rate

Per-stage scores — Awareness, Consideration, Decision

Add / edit / delete prompts inline

Direct link to full AI Visibility report

Keyword Intelligence

Keywords with
context and source

intelligentAEO doesn't just track keywords — it mines them from page content, enriches them with ranking data, and clusters them by topic and intent. Every keyword carries a source tag so you always know where it came from.

Step 1 — Discovery

Multi-source keyword mining

Keywords are discovered from four sources, each tagged. Gemini extracts target keywords from crawl data. GSC surfaces real search queries with impression and click data. DataForSEO/SEMrush adds volume and competition. Manual entry covers strategic terms not yet in the data.

crawl gsc semrush manual

Step 2 — Classification

Intent tagging on every keyword

Every keyword is classified by intent — informational, commercial, transactional, or navigational. The keyword store tracks target URL, section, cluster, status, and source alongside the keyword itself.

Step 3 — Clustering

Topic-based grouping via Gemini

The Keyword Cluster agent batches all unclustered keywords and asks Gemini to assign consistent topic-based cluster names. Existing clusters are passed as context so new batches maintain naming consistency. A --normalize pass merges similar clusters and title-cases names.

Step 4 — Rankings

Live SERP positions at scale

Active keywords run through DataForSEO for current rank positions. Results feed back into the cluster view — giving a ranking snapshot per topic group, not just per keyword. Cost: ~$0.0006 per keyword.

Sample cluster output

Jewelry Care & Cleaning

19 kws · informational

Charm Bracelets

18 kws · commercial

Gifts For Her

16 kws · transactional

Lab-Grown Diamonds

14 kws · commercial

Personalized Jewelry

13 kws · commercial

Competitor Comparison

7 kws · commercial

Quality Signals

8 kws · commercial

Gifting Occasions

8 kws · commercial

Tennis Bracelets

7 kws · commercial

Promise Rings

7 kws · commercial

38 clusters · 252 keywords · 4 sources

AI Visibility

Measure how AI
sees your brand

As more shoppers use AI assistants to discover, compare, and decide — visibility in AI answers matters as much as search rankings. intelligentAEO measures brand presence across the full purchase funnel with structured, repeatable prompt testing.

Awareness

Discovery queries

Tests whether the brand appears when shoppers are exploring a category or need — without naming the brand.

"what jewelry makes a good gift for a girlfriend's birthday"

Consideration

Comparison queries

Tests how the brand is represented in direct comparisons — quality, value, and positioning against competitors.

"Pandora vs Tiffany charm bracelets which is better quality"

Decision

Purchase enablers

Tests whether AI can surface accurate policy and purchase information — returns, shipping, warranty, care.

"does Pandora offer free returns on online orders"

—

Awareness score

—

Consideration score

—

Decision score

—

Overall / 100

What the scores reveal

Each prompt is scored 0–100 for brand visibility, with sentiment, citation type, and competitor presence extracted from the AI response. Patterns across prompts reveal where a brand wins by being known, and where it loses because its content is inaccessible, unstructured, or missing the signals AI models rely on to form recommendations. The gap between Decision-stage strength and Awareness-stage absence is almost always an infrastructure problem — not a brand problem.

Intelligence-drivenAEO at scale

Bottom-up analysis,four levels deep

Every capabilityis a registered agent

One dashboard.Every site.

Keywords withcontext and source

Measure how AIsees your brand

Connected to thetools that matter

Intelligence-driven
AEO at scale

Bottom-up analysis,
four levels deep

Every capability
is a registered agent

One dashboard.
Every site.

Keywords with
context and source

Measure how AI
sees your brand

Connected to the
tools that matter