Tools that bridge search science and AI retrieval
Three kinds of tool here: free interactive utilities you can run right now, no sign-up; the three tools I built at Wellows, grounded in Google patent research and used by 1,647 people worldwide with zero paid spend, led by the Query Fan-Out Generator that Search Engine Land picked to demonstrate query fan-out; and the open-source SEO projects on my GitHub.
What tools has Khadija Zaman built?
Khadija Zaman built three SEO tools at Wellows grounded in Google patent research, the Query Fan-Out Generator, the LLM Query Builder and the AI Humanizer, used by 1,647 practitioners with no paid acquisition. She also publishes four free in-browser tools on this site and six open-source SEO projects on GitHub.
Try these free, no email required
AEO Citeability Checker
Paste your content and score how ready it is to be cited by ChatGPT, Gemini, and Perplexity, with specific fixes.
Check my content →SERP Snippet Preview
Approximate pixel-width preview of your title and meta description in Google, with truncation warnings. Google may still rewrite or truncate differently.
Preview my snippet →Query Fan-Out Explorer
Enter a seed query and see it expanded into 44 template sub-queries grouped by fan-out type, as a planning aid.
Expand a query →Schema JSON-LD Generator
Fill a short form and copy ready-to-paste Article, FAQ or Person structured data that helps engines parse the page (FAQ rich results are limited to government and health sites).
Generate schema →Built at Wellows for practitioners, used by 1,647
1,647 is the three tools combined (829 + 149 + 669), from my own tracking while I was building and measuring them. These are not the tools’ lifetime totals: Wellows’ live counter for the Query Fan-Out Generator reads 6,849 in the Search Engine Land feature, because the tool kept acquiring users long after I stopped counting.
Query Fan-Out Generator
Built on Google's query fan-out patent research: the sub-query expansion AI Mode and AI Overviews run before they answer. Enter one seed query and it returns 40+ semantically distinct variants, each classified by intent (informational, commercial, comparative, navigational) and relevance, with CSV export for content planning. Use the export to see which variants your pages don't cover yet. The free Query Fan-Out Explorer above is a lightweight, on-site preview of this tool.
- 829 active users in my tracking period
- 6,849 lifetime users on Wellows’ live counter
- 5.1% CTR from organic search
Featured in Search Engine Land. Its guide to topic clusters for AI search uses this tool to demonstrate the method: “There are multiple query fan-out tools. For this tutorial, we’ll use Wellows.” The article credits Wellows, where I built it; the COO’s public post on the experiment behind it names me. Read the Search Engine Land guide →
Listed in Search Engine Land’s query fan-out tools roundup alongside the category’s other software. See the roundup →
LLM Query Builder
Generates the natural-language queries a buyer would actually type into ChatGPT, Perplexity, or Gemini about your category, spread across personas and four intent types (informational, navigational, commercial, transactional). The hosted builder returns 40 at a time; the open-source engine behind it, on GitHub below, can generate 100+. Built for practitioners who need to work with AI retrieval systematically.
Launched on Product Hunt. Masab Gadit’s launch post credits me for the build. Read Masab Gadit’s Product Hunt post →
AI Humanizer
Transforms AI-generated drafts into content that reads naturally and holds topical depth. Paste a draft, set tone and strength, and compare before-and-after stats. Its GA4 engagement rate (engaged sessions as a share of all sessions, where an engaged session lasts over 10 seconds, has a key event, or has two or more page views) ran between 60% and 72% across the months I tracked it.
Open-source SEO and AI-search tooling
Everything below is public on github.com/KhadijaZaman. Several of these codebases power the production tools above.
LLM AI Search Query Builder
The engine behind the LLM Query Builder. A 9-step pipeline reads a site’s homepage, feature, and pricing pages, extracts products, use cases, audiences, and plans, maps them to personas and pain points, and can generate 100+ natural-language queries across the customer journey, then de-duplicates and strips competitor brand names.
LLM AI Search Query Builder on GitHub →Domain Authority Explorer
A free, open-data alternative to proprietary authority scores. Ingests the Common Crawl Web Graph (about 121 million domains) and returns two 0–100 metrics: an Authority Score from harmonic centrality and a Popularity Score from PageRank, with monthly history back to January 2018 and a corpus audit of how often a domain is captured.
Domain Authority Explorer on GitHub →Citation Auditor
Full-stack app that scrapes a set of pages, scores each one for the signals AI answer engines cite (the same criteria as the free Citeability Checker above, run at scale), and exports the results as a spreadsheet so a content team can work through them.
Citation Auditor on GitHub →Internal Link System
A monorepo for AI-assisted internal linking: a typed API spec and Zod schemas, a database layer, a React client, and a Claude batch integration that reads your pages and proposes contextually relevant internal links instead of keyword-matched ones.
Internal Link System on GitHub →Wellows MCP Server
A production Model Context Protocol server that lets Claude and any MCP client check a domain’s citation score inside Google AI Overviews without opening a dashboard. Four tools: extract domain entities, generate 40 intent-driven queries, scan live AI Overviews for citations, and benchmark the resulting score with next actions.
Wellows MCP Server on GitHub →Custom Semantic Search Engine
Where the retrieval interest started: a semantic document search built from scratch with Word2Vec embeddings, cosine-similarity ranking, and BART summarisation, packaged as a Colab notebook with explanations and exercises. Useful if you want to see how vector retrieval works under the hood before optimising for it.
Custom Semantic Search Engine on GitHub →Authority signals
I built three tools at Wellows to connect Google's patent research to AI-retrieval content work, and open-sourced the engines and the surrounding tooling. They onboarded 1,647 practitioners across the UK, US, and EU with zero paid spend, and Search Engine Land picked up the Query Fan-Out Generator twice.
Grounded in research
Every tool starts from primary source material: Google's query fan-out patents and how retrieval systems actually score and select content, not recycled best-practice blog posts. Search Engine Land used the Query Fan-Out Generator to demonstrate the method.
Distributed by usefulness
1,647 users, acquired with zero paid spend. The tools spread because they solve a real problem.
Tied to outcomes
They are part of the system that grew Wellows’ LLM referral sessions 9× in 8 months and lifted GSC impressions from 680K to 3.78M.