Entity Intelligence Pipeline
30+ live collectors across 8 signal categories — flowing through 6 stages and 5 continuous learning loops to one canonical entity.
What feeds the pipeline
Every signal category is wired to live collectors. The pipeline begins by running these in parallel against the canonical entity.
AI engines & answer surfaces
How generative engines describe, recommend and cite your entity.
Identity & structural authority
Canonical identity and structural credibility across registries and knowledge bases.
Reputation & reviews
Customer-facing reputation networks where perception is recorded directly.
Community & social
Where people actually discuss you — developers, traders, fans, critics.
News
Press coverage and editorial framing across global news APIs.
Search & technical
Search visibility, AI Overviews and the technical health that backs both.
Geographic
Real-world location signals for places, venues and physical entities.
Sentiment NLP
Multiple independent NLP engines for triangulated sentiment.
Live web index sweep
Before the main pipeline runs, a cross-surface discovery sweep indexes how the entity currently appears across live web surfaces — organic search results, AI Overviews, knowledge panels, news, places, academic indexes, and more. Results are routed to the correct signal surface, written as provenance-rich observations, and fed into drift detection.
Output
Provenance-rich observations → Drift events → Surface coverage score
The private intelligence engine. Entidex collects signals from 15+ sources, resolves entities to canonical identities, builds evidence graphs, and runs 5 continuous learning loops to produce deep entity intelligence.
The public Explore Entidex. Entidex surfaces visibility scores, sentiment signals, recommendation presence, and narrative themes — all derived from Entidex intelligence.
Continuous Learning Engine
Entidex5 autonomous loops driving self-improving intelligence — Entidex learns from every observation cycle
The 30 active collectors
Every collector below is wired up and running today. This is the canonical list — it powers homepage stat strips, pricing tier coverage, and Milo's evidence drilldowns.
| Category | Collector | Status |
|---|---|---|
| AI engines & answer surfaces | ChatGPT | Live |
| AI engines & answer surfaces | Gemini | Live |
| AI engines & answer surfaces | Claude | Live |
| AI engines & answer surfaces | Perplexity | Live |
| AI engines & answer surfaces | Google AI Overviews | Live |
| Identity & structural authority | Wikipedia | Live |
| Identity & structural authority | Wikidata | Live |
| Identity & structural authority | Google Knowledge Graph | Live |
| Identity & structural authority | Companies House | Live |
| Identity & structural authority | Charity Commission | Live |
| Identity & structural authority | WHOIS | Live |
| Reputation & reviews | Trustpilot | Live |
| Reputation & reviews | Google Maps Reviews | Live |
| Community & social | Live | |
| Community & social | Bluesky | Live |
| Community & social | Hacker News | Live |
| Community & social | GitHub | Live |
| Community & social | YouTube | Live |
| Community & social | Podcasts | Live |
| News | Global news | Live |
| News | International news | Live |
| News | Global news index | Live |
| Search & technical | Search index | Live |
| Search & technical | SEO authority | Live |
| Search & technical | Web Vitals | Live |
| Search & technical | Schema.org check | Live |
| Geographic | OpenStreetMap | Live |
| Sentiment NLP | NLP analysis | Live |
| Sentiment NLP | Linguistic NLP | Live |
| Sentiment NLP | Sentiment engine | Live |
Live web index sweeps and cross-source drift detection are now live \u2014 see Stage 0.5 above. Roadmap collectors (Bluesky firehose, event detection feed) are tracked in methodology.