How I Made My Site Visible to AI Assistants
An argument by Rik Avalos
At Trinzik, AI visibility is a big part of what we do. Companies want to know what ChatGPT, Perplexity, and Google's AI answers say about them, and why those answers are often thin or wrong.
Recently I realized I'd never run that test on my own name.
What AI saw
The answer wasn't great. My site lacked authority. It had no certifications. It didn't read as secure. And for the most part, it wasn't visible at all.
That stung a little. I've spent twenty years around companies like Tesla, Google, and Neuralink, and none of it was coming through. The information existed. It just wasn't in a form a machine could trust.
What I changed
First, I told the machines who I am. There are other people named Rik Avalos, so the site now carries structured data, a behind-the-scenes description that search engines and AI tools read, connecting my work at Tesla, Neuralink, and Trinzik, my band, and my press mentions into one profile. I also created a Wikidata entry, which is one of the sources AI systems lean on to tell people apart.
Second, I backed every claim with a source. I'm named on pages 139 and 162 to 163 of Tim Higgins's Power Play, the book about Tesla's early years, and The Washington Post quoted me in its review of it. Those are now cited with page numbers and links. I only added things I could verify. When a book summary site turned up that just repeated the Power Play material, I left it out, because it didn't add a new source.
Third, I made the site show it could be trusted. It now carries a TrustedSite security certification. That meant pasting a script onto every page, which is exactly what a phishing attack looks like, so I verified the script came from TrustedSite before adding it.
Fourth, I wrote for AI directly. The site has an llms.txt file, a plain-text summary at a standard address that AI systems can read without digging through the design. It updates itself whenever I publish something new, including this essay.
Testing it
I ran the site through Google's Rich Results Test, which checks whether that structured data actually works. It found two errors I never would have caught by looking at the page. Both are fixed, and every page now passes.
What changed
The authoritative links about my Tesla years, the book review that mentions my work there, and the certification all give an AI system something to check my story against. When I check now, it's the opposite of where I started. The site reads as authoritative, certified, secure, and visible.
It isn't finished. My Spotify, Apple Music, and YouTube profiles still don't link back to the site, and those links help AI systems confirm that the musician and the recruiter are the same person. That's next on the list.
I didn't do this alone
I'm not an engineer, and I didn't pretend to be one here. Bob Michaels, my co-founder at Trinzik, set up the infrastructure this site runs on, and he brings thirty years as a developer. I did the work through Claude Code, but his experience was behind every request I made.
That's the honest version of "I built this myself." I built it with a strong foundation under me.
Why I started with my own name
It was personal, and it was business, and both mattered a lot.
Personally, I didn't want the first thing an AI assistant said about me to be nothing. Professionally, this is the work Trinzik does for companies, and I wanted to know what it takes when it's my own name on the line. Now I can tell a client exactly what the first test felt like, because I took it.