Does your resume sound like AI wrote it?

A 0–100 measure of how much your resume bullets read as AI-generated. It uses the same scanner Oliros runs on every kit we generate: a curated list of ~80 phrases recruiters cite as AI tells (results-driven, leveraged, spearheaded, cross-functional collaboration to drive strategic initiatives, etc.) plus a burstiness check on sentence rhythm. Higher score = reads more like you, less like AI.

62% of employers say AI-generated resumes without customization are more likely to be rejected, and 49% would dismiss resumes they merely suspect are AI. The rejection signal isn't actually 'used AI' — it's specific phrases recruiters associate with it. Avoiding the phrases (without losing your voice) is the defensible move.

Nowhere. The scoring runs entirely in your browser using a pure JavaScript function. Your bullets are not uploaded, not logged, not stored. You can verify by opening DevTools → Network — zero requests fire when you score.

GPTZero and Originality.ai try to detect whether AI wrote your text using stylometric analysis (perplexity + burstiness). They have 15–33% false positive rates, and 23% for non-native English speakers. The Oliros AI-Tell Score is different: it specifically scores against the canonical phrases recruiters reject for — the actual signal that matters in hiring, with zero false-positive risk on non-native English.

Yes — that's what Surgical Tailoring does. We rewrite flagged bullets with minimum edits (preserving your verbs, rhythm, and specifics), inject your real JD keywords, and reject any output that re-trips the AI-Tell scanner. Free to start, three variants per bullet, 'use original' is always an option.