Cailyx · Our engine
Know how AI recommends you. Then change it.
Cailyx is the engine Rothenhall runs every engagement on. It measures how AI answer engines see your brand across the queries buyers actually ask, builds the entities, content, and citations that move that view, and reports the result as a rate you can track, never a rank.
Access is set up by Rothenhall with your workspace, on onboarding. No card on this page.
The Cailyx client workspace, shown with a fictional sample company. Figures are illustrative.
Measured across the seven surfaces buyers ask
- ChatGPT
- Perplexity
- Google AI Overviews
- Google AI Mode
- Gemini
- Google Search
- Copilot
The problem
Buyers shortlist inside an AI answer now. Most companies cannot see theirs.
A buyer asks an AI who to consider, and the answer names a handful of companies. That list forms before anyone opens your site.
You cannot manage what you cannot see, and AI visibility is almost impossible to see. Teams fall back on a single screenshot of one chat, a gut feel, or a dashboard that reports a score and stops there. Nobody shows you how often you are actually named, which competitors the model prefers, or what to build to change it.
There is a measured way.
Inside Cailyx
The workspace your team gets, screen by screen.
Not a slide of promises. These are the surfaces Cailyx puts in front of you: the measurement, the competitors, the technical evidence, and the fixes that prove themselves.
The answer comes first: how often AI names you, how healthy your site is, which fixes are moving, and what needs you today.
- 1How often AI engines name you, with the rate behind it
- 2Technical health out of 100, with the trend
- 3Fix Plan progress, and decisions waiting on you
- 4A short list of what needs attention now
The real Cailyx workspace, shown with a fictional sample company. Figures are illustrative.
Capabilities
One engine for the whole AI-visibility loop.
Make sure the AI can reach you
Cailyx fetches your site the way crawlers do, probes robots.txt and the CDN with real AI user-agents, and flags the silent blocks and JavaScript-render walls that turn an assistant away before it reads a word.
Give models a name they can resolve
It audits your structured data, checks your entity across the web, and disambiguates you from a same-named company, so an assistant knows it is reading you and not your rival.
Measure how you show up, as a rate
Against a query set you own, Cailyx runs each prompt across the answer engines five or more times and reports your mention rate, citation rate, and share of voice. A distribution, not one lucky answer.
Show the shortlist you are missing
It names the competitors the models recommend in your category and where they beat you, so you know exactly which answers and which lists to win.
Build the fix, not just the finding
Agentic workflows draft the answer-shaped content, the internal and third-party links, and the structured entities that raise citation share. A Rothenhall operator reviews and ships it, so the work gets done.
Simulate the buyer before the buyer does
Cailyx runs synthetic personas through realistic journeys, and a panel of review agents stress-tests the plan, so you find the gaps in the answer before a customer does.
Track the score over time and alert you
Your AI Visibility Score, its five dimensions, and your share of voice sit on a monitored trend. When it moves, up or down, Cailyx tells you.
How it runs
From invisible to measured to moved, in five steps.
Intake
You set the category, the competitors, and the buyer questions that matter.
Query set v1
Measure
Cailyx samples every engine, repeatedly, and records how you are described, cited, and ranked.
Runs on 7 surfaces
Diagnose
The evidence rolls into an AI Visibility Score and the reproducible reasons behind it.
Score + reasons
Build
Agents and your operator produce the fixes: schema, content, citations, links.
Fix Plan
Watch
Cailyx monitors the shift, so growth stays provable and nothing quietly regresses.
Trend + alerts
Why the number holds up
Measured like research, not like a screenshot.
- Prompts run in the markets you serveAI answers vary by location, not a global default.
- A named, versioned query set that you ownNot a preset list chosen for you.
- Distributions and share of voice, never a claimed rankThere is no “first” in a synthesized answer.
- Every score traces to the runs and checksEvidence-linked, reproducible, auditable.
- Probabilistic honestyCailyx reports likelihood, not promises.
A visibility tool you cannot reproduce is a mood. Cailyx ships the method to reproduce every claim.
What we will not tell you
No one can guarantee an AI answer. We do not pretend to.
Can you get us to rank first in ChatGPT?
There is no “first” in a synthesized answer. We move the rate at which you are named and cited, and we show you the evidence.
Do you just hand us a report?
No. The agents and your operator build the fixes, and you keep a workspace to see them and act.
Is this a login we sign up for ourselves?
Not anonymous. Rothenhall sets up your workspace on onboarding and runs the deeper work behind the scenes.
A Cailyx MCP for our own AI stack?
On the roadmap. Join the list below to be first when it opens.
Proof, not claims
Engagements run on the engine, baselines captured at kickoff.
55 → 62
DayOne Technologies, +7 points on the AI Visibility Score in the first remediation phase, with 3 new qualified leads.
43 / 100
Napkin, a baseline the owner could not have seen alone, after the engine surfaced a silent CDN block on AI crawlers.
90k
AI answers across 15+ industries in Rothenhall’s field study, the ground under the rubric and the method.
Also from Rothenhall · Free tools
Two things we built for founders. Use them free, no account.
Your buyers are asking AI who to trust. See exactly what it says about you.
Request access and Rothenhall sets up your Cailyx workspace with your first measured score.
Cailyx MCP
Be first to the Cailyx MCP.
A Cailyx MCP, so it can plug into your own AI workflows, is on the roadmap. Join the list and we will reach out when early access opens.