Founder
Ekeleme David Kelechi
KellyxyHub · Founder of SiteNexis
AI Visibility Researcher, educator, and the person who asked why excellent content was consistently invisible to AI systems — then built the infrastructure to answer that question.
Founder Story
The gap that became the platform
I spent years teaching, building websites, and helping businesses improve their digital presence. During that time, I developed a detailed working understanding of how search engines operate, how content gets ranked, and what the gap between publishing and being found actually consists of.
Then AI search changed the question.
As ChatGPT, Gemini, Claude, Perplexity, and Google AI Overviews became significant discovery surfaces, I started noticing a consistent pattern: content that ranked well in traditional search was often completely absent from AI-generated responses on the same topics. The SEO was working — the rankings were there. But the AI systems were not citing those pages.
No existing tool could explain why. Traditional SEO audits reported healthy scores. Schema validators showed no errors. The technical foundation was sound. But the AI systems kept ignoring the content.
I spent months researching the mechanism: how AI systems chunk content into semantic units, how they evaluate entity clarity, how they apply trust signals at the domain level before evaluating individual pages, and how citation eligibility filtering works as a final screen that well-structured content can still fail. That research became SiteNexis.
Why SiteNexis Exists
Traditional SEO tools measure rankings.
No one was measuring machine trust.
The tools available to website owners in 2024 were all built for the traditional search model: a user inputs a query, a search engine returns a ranked list of results, the user clicks. Every metric in the standard stack — rankings, impressions, CTR — measures performance in that click-mediated model.
AI search systems do not work that way. They receive a query, generate an answer using sources they select based on their own trust and citation logic, and return a synthesised response. The user may not see a ranked list. They may not click through to any source. The brand that gets cited in the AI response receives the visibility benefit. The brand that does not is invisible to that user, regardless of how highly it ranks in traditional search.
SiteNexis exists to close that measurement gap. It models the complete AI retrieval and trust pipeline — from technical crawl accessibility through entity clarity, machine trust signal consistency, retrieval simulation, temporal authority, and recommendation surface coverage — and makes every dimension measurable, explainable, and improvable.
4
Intelligence Layers
16
Analysis Agents
12
Scored Dimensions
Professional Journey
From the classroom to the intelligence layer
Foundation
Educator & Developer
Spent years teaching digital skills and building web projects for businesses — developing a detailed understanding of how the web works from both the human and machine perspective.
Observation
AI Search Emergence
Tracked the rise of ChatGPT, Gemini, Claude, Perplexity, and Google AI Overviews. Observed a consistent pattern: well-ranked content was often invisible in AI-generated responses.
Research
AI Retrieval Investigation
Spent months researching why the gap existed — studying how AI systems chunk, embed, evaluate, and select content. Identified entity clarity, trust signals, and machine readability as the primary determinants.
Building
SiteNexis
Built SiteNexis to systematically measure and explain AI visibility — a four-layer intelligence stack that models everything from technical crawl accessibility to machine trust formation and recommendation surface coverage.
Research Philosophy
How I approach every problem
Mechanism before conclusion
Every finding starts with the mechanism: why does this pattern exist? What specific property of the AI retrieval pipeline produces this outcome? Conclusions without mechanisms are difficult to act on.
Deterministic over probabilistic
Where measurement is possible, it should be reproducible. The same content should produce the same score. Stochastic evaluation creates doubt that prevents trust in the findings.
Explainability as a hard requirement
A score without a sub-score breakdown is a black box. A deduction without a named cause cannot be fixed. Every metric in SiteNexis traces to a specific, actionable reason.
Semantic quality over signal gaming
AI visibility that is manufactured without corresponding content quality is fragile — it decays as AI systems improve their detection of inauthentic signals. Durable visibility requires genuine quality.
Expertise
Areas of deep focus
AI Retrieval Systems
Modeling how large language models and embedding-based retrieval systems process, chunk, and evaluate web content for citation eligibility.
Machine Trust Intelligence
Researching the trust signal architecture that AI systems use to determine source credibility — entity consistency, schema alignment, external validation.
Entity SEO
Advancing the practice of entity-first content strategy: structuring web content as knowledge graph contributions rather than keyword collections.
Prompt Engineering
Designing structured AI prompts for content evaluation, entity extraction, contradiction detection, and retrieval simulation at scale.
Web Development
Full-stack development with Next.js, TypeScript, and modern web infrastructure — from architecture design to production deployment.
AI Visibility Measurement
Building scoring systems that produce deterministic, explainable AI visibility metrics — every deduction maps to a named, actionable issue.
Vision
The future of AI visibility
The transition from keyword-based search to AI-mediated discovery is the largest structural change in how web content is found since the invention of PageRank. It is not a complete replacement — traditional search ranking will remain significant for years. It is an expansion: a new layer of the discovery stack that operates on different principles and requires different measurement.
The organisations that understand this transition early will build structural advantages that compound: better entity clarity produces better AI citation rates, which produces stronger brand recognition in AI-generated responses, which produces higher branded search volume, which produces more external validation signals, which produces better entity clarity. The loop is self-reinforcing.
The future I am building toward is one where every brand — regardless of size — can access the same intelligence layer about their AI visibility that currently only the most sophisticated AI-native companies have. That means not just scores, but explanations. Not just diagnosis, but prioritised action. Not just a snapshot, but a temporal model that tracks how machine trust is growing, stable, or declining.
That is what SiteNexis is. And we are only at the beginning of what that intelligence layer can produce.
Thought Leadership
Writing, research, and speaking
My research and writing on AI visibility, machine trust intelligence, and the evolution of search is published on the SiteNexis blog. I write from direct observation of AI retrieval behaviour — not from second-hand summaries of industry trends.
Mission
“Every brand that produces genuine value deserves to be understood and trusted by AI systems — not just by human readers. Building that understanding is not a luxury. It is the next infrastructure layer of the web.”
— Ekeleme David Kelechi
See how SiteNexis works
Run a free AI visibility audit on any domain. Understand exactly where and why AI systems are or are not citing your content.