Waikay’s Entity Map Generator
Waikay builds a complete, interactive map of every entity and relationship on your site. See it visually, edit it directly, and watch it evolve with your content – then publish it as a structured file AI systems can actually use.
Not a file. A living map.
AI systems retrieve content from your site by reading raw HTML and chunking it into passages – with no awareness of which entity a passage is about, how your concepts relate, or who published it. Your expertise is fragmented. Your brand attribution gets lost. The relationships between your concepts are invisible.
Waikay’s Entity Map Generator fixes this. We crawl your site, extract every entity and relationship, and give you a visual, interactive knowledge graph you can actively work with – not just a file to publish and forget.
Create
Built automatically from your site. What would take weeks of manual JSON editing is ready in minutes.
Visualise
Your entire knowledge graph rendered as an interactive map. See every concept, every relationship, every gap at a glance.
Edit
Fix gaps, add relationships, update canonical names – directly inside Waikay. No JSON, no files, no technical overhead.
Evolve
As you publish new content Waikay detects new entities and flags them for review. Your map stays current automatically.
Everything inside Waikay’s Entity Map Generator
Interactive Visual Graph
Explore your entire knowledge architecture as a live, interactive graph. Click into any entity, trace relationships, and understand exactly how AI sees your content.
Gap and Error Detection
Waikay surfaces thin coverage, missing relationships, and disambiguation issues with clear guidance on how to resolve each one.
In-Tool Editing
Add entities, define relationships, fix canonical names directly inside Waikay. No code, no files, no friction. Changes update in real time.
Automatic File Publishing
Waikay generates your entitymap.json and entitymap.html files with publisher attribution correctly declared on every chunk – ready for all AI crawlers.
Continuous Monitoring
As you publish new content Waikay detects new entities and relationships and flags them for review. Your EntityMap stays current automatically – no set-and-forget decay.
Three things going wrong in AI retrieval right now
Disambiguation Loss
“AI SOV”, “AI Share of Voice”, “artificial intelligence share of voice” – to a page-level retriever these are three separate signals. Your expertise dilutes across fragments instead of concentrating on a single recognisable entity. EntityMap resolves them all to one canonical concept.
Attribution Loss
Your content is used in an AI answer. The URL appears as a footnote. Your company name never appears. Publisher identity does not survive aggregation. EntityMap declares attribution on every chunk in a field designed to survive redistribution.
Reasoning Loss
The relationships between your concepts are scattered across pages. AI reconstructs them probabilistically and hedges with “likely” and “probably” where your content is definitive. EntityMap makes the logic explicit, typed, and readable.
EntityMap directly addresses retrieval-time problems: disambiguation, attribution, and reasoning in live RAG pipelines. It does not fix wrong associations already baked into a model’s training weights. The ghost citation problem has a retrieval half (fixable now with EntityMap) and a training data half (not fixable post-training). EntityMap solves the retrieval half – the part within a publisher’s direct control.
EntityMap is also a free open standard
Waikay’s Entity Map Generator is a paid tool that builds and manages your knowledge graph for you. But the EntityMap standard itself is free and open for anyone to implement manually – published openly by the Waikay founders.
A file at yourdomain.com/entitymap.json tells any AI consumer: here are the concepts this site covers, here is how we define them, here is the best evidence for each, and here is how they relate to each other.
Implementing it manually – writing and maintaining the JSON, keeping it current, ensuring every relationship is correct – is a significant ongoing effort. That is exactly the problem Waikay’s Entity Map Generator solves.
In 2005, sitemap.xml solved the discovery problem for search engines. Within two years every major CMS generated one automatically. EntityMap is the equivalent structured layer for AI retrieval – and Waikay is the tool that makes it as effortless to maintain as a sitemap.
Three steps to implement EntityMap on your site:
1. Publish the JSON file at the root of your domain
2. Publish the companion HTML file at the same root
3. Reference the HTML file in your site footer
Add this to every page so AI crawlers can find it from anywhere on your site:
You do not need Waikay to use EntityMap. The standard is free and open – you can implement it manually today. Waikay’s feature is for organisations that want the full benefit without the maintenance burden: built, visualised, edited, and kept current automatically.
Frequently Asked Questions
The Waikay’s Entity Map Generator feature is coming soon. Join the waitlist to get early access and be notified at launch. The open EntityMap standard is available now and can be implemented manually while you wait.
No. Waikay’s Entity Map Generator is entirely no-code. You interact with your knowledge graph through a visual interface and Waikay handles file generation and publishing automatically.
Yes. The EntityMap standard is open and free. Waikay’s feature is a paid product that handles creation, visualisation, editing, and maintenance automatically. The standard and the tool are separate – you do not need Waikay to implement EntityMap, but Waikay makes it significantly more powerful, accurate, and maintainable.
Partially. Ghost citations have a retrieval component – where your content is used but your name does not appear – and a training data component – where associations are baked into model weights. EntityMap fixes the retrieval half, which is the part any publisher can actually control. The training half requires retraining the model.
Schema.org annotates individual pages with facts about the content on that page. EntityMap operates at the site level – declaring the full set of concepts a site covers, how they relate to each other, and where the best evidence lives. It is a site-wide knowledge graph, not page-level annotation. The two are complementary, not competing.
The first living Entity Map Generator.
See your knowledge graph. Edit it directly. Watch it evolve with your content. Built on the open standard, powered by Waikay.
Join the Waitlist