Intelligence-driven
AEO at scale

A custom Python pipeline that transforms crawl data into prioritized, AI-powered AEO and SEO strategy. It measures how AI sees your brand, surfaces what's blocking visibility, and builds the keyword intelligence to act on it.

See how it works Explore the agents →
10+
Specialized agents
4×
Analysis levels
28+
Crawl signals captured
3
AI funnel stages tracked
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.

01
Crawl
Playwright: 28+ fields, AI readiness scoring, JS render ratio.
02
Technical Audit
30+ rules-based checks. AI Readiness 0–100. No API calls.
03
Keywords
Gemini extraction → clustering → DataForSEO rankings.
04
Content Briefs
Gemini: keyword, heading outline, rank-aware guidance.
05
Section Insights
One Gemini call per section. Funnel-stage aware synthesis.
06
Cross-Section
Site-wide patterns, priority actions, executive summary.
07
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. Each declares its 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 straightforward 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 mines keywords 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 a structural infrastructure problem.
Integrations

Connected to the
tools that matter

intelligentAEO routes through the best available data source for each task, with graceful fallbacks and 7-day caching to minimize API costs.

Gemini 2.5 Flash
Live — analysis engine
Playwright
Live — JS crawl engine
DataForSEO
Live — SERP + rankings
Google Search Console
Live — per-site credentials
Google Analytics 4
Live — per-site credentials
FastAPI + HTMX
Live — command center
SEMrush
Planned — volume + KD
Bing Webmaster Tools
Live — crawl + index data
Reddit API
Beta — brand mentions
Case Studies

The platform
in practice

Real audits run on live sites. Every finding, ranking, and score below was produced by intelligentAEO running on real data.

Jewelry E-Commerce · March 2026 · us.pandora.net
Pandora — Full-Funnel Audit
AI visibility across three purchase stages, keyword intelligence across 252 tracked terms, onsite readiness scoring by section, and competitive gap analysis. Six previously untracked competitors surfaced from SERP data.
87/100
decision AI score
75pt
funnel gap
252
keywords tracked
6
new competitors found