Framer AI AgentsDesign and publish professional sites with AI
Promoted
Centauri
Analyze content beyond AI discoverability spam
11 followers
Analyze content beyond AI discoverability spam
11 followers
Centauri is an SEO and AI discoverability analyzer tool built for the world after AI spam, where originality, accuracy, authority, readability, and structure matter more than shortcuts.
Hi Product Hunt 👋
I'm Pankaj, one of the creators of Centauri.
AI discoverability has turned into a spam era, and honestly, it's a bit exhausting.
Every week, there's some new hack going around. - Add this file. Rewrite it this way. Spin up fifty pages instead of one. Format it for whatever model's flavour of the month.
And yeah, some of it probably works right now because AI search is still figuring itself out. Different tools are testing different ways to retrieve and rank content, which is exactly why this phase feels so gameable. That's not going to last forever.
Once things settle, the shortcuts stop working. What's left is just... good content. Original, accurate, actually useful, easy to read, well put together.
It doesn't sound too exciting, but it's true. That's what we built Centauri for.
So what does it do?
It scores your content before you hit publish, not after you're staring at the analytics wondering why nothing ranked. It checks originality, accuracy, authority, readability, structure, and 30+ factors in total, and then tells you exactly what to fix, down to the sentence level.
Basically, instead of hoping your content ranks, you know exactly what to change.
On top of that, it lives where your team already works. Pull content in from Google Docs, get the feedback back as comments in the same doc. No new tool, no copy-pasting between five tabs.
One thing we're clear on: Centauri isn't trying to be a one-click content machine. We don't think that's the answer. AI can speed things up, sure, but someone still needs to make the final call before something goes out.
Think of Centauri as the quality check, not the replacement. We're rolling out access in batches. You can start a free trial today, no credit card needed.
If you try it, run it on one article you're about to publish. Tell us what actually helped, what felt confusing, and what you wish it did.
I'll be around, ask me anything, anytime.
How does Centauri actually measure originality and authority versus just scraping a few public metrics? Curious whether the scoring is based on third-party signals or something you built on your end.
@batuhana92994 Originality, authority, and credibility are interrelated metrics in Centauri. And these metrics are judged and weighted upon on the sentence level, section level, and overall content level. The weighting is also contextual. We use a combination of other smaller metrics to deduce these scores.
Because of this, one wrong fact in a content piece may not affect the credibility of the content piece by much, but sometimes one single wrong fact could bring the credibility of the entire content piece down. It also varies by industry; for example, a wrong healthcare fact could discredit your entire content piece in the industry.
Similarly, originality and readability are a combination of other smaller metrics talking to each other. It was a tough nut to crack, but we had to because in my experience, if a content piece is not original, not credible, and does not read well, it has a lower chance of getting discovered by search engines and AI.
How does it actually judge "originality" and "authority" in practice, since those feel pretty subjective without a clear scoring rubric?
@gkhan2f3a We don’t score originality and authority as isolated subjective metrics. For originality, Centauri checks whether the content is only repeating already common information or adding something less obvious, specific, or owned by the publisher. That could come from first-party data, internal experience, examples, or a clear point of view.
But we also understand that if something is new but not credible, it does not help. So Centauri also checks whether the content supports claims properly, either through credible external sources or through clear first-party evidence.
Authority is judged by how convincingly the content is written and structured at the sentence, section, and full-article level. Does it explain claims clearly? Does it show depth? Does it avoid vague or unsupported statements?
Credibility sits underneath both originality and authority. That’s also how AI systems are likely to judge content over time (search engines are already doing it).
The readability and originality scores actually match how I'd describe my own writing, which surprised me. Useful for catching thin content before publishing.
@songlmtgi That’s great to hear. We wanted the scores to reflect how people actually read the content piece.