HackerRank open sourced its ATS. My resume scored 90/100. Oh wait 74. No – 88

	HackerRank open sourced its ATS. My resume scored 90/100. Oh wait 74. No – 88(danunparsed.com)
	1012 points by sambellll 54 days ago \| 432 comments
	tl;dr: HackerRank's open-sourced ATS (hiring-agent) produces wildly inconsistent resume scores—the same resume scored anywhere from 66 to 99 across 100 runs, meaning candidates can fail an 85-point cutoff 65% of the time purely by luck. The author traces this to LLM non-determinism on subjective judgments (projects vary hugely, while checklist-style skills stay consistent) and a two-line "experience" prompt with no rubric that awards 25/25 to everyone from interns to principal engineers. The takeaway: LLMs are fine for parsing resumes but shouldn't be making qualitative scoring calls that decide who gets filtered out.
	HN Discussion: ↑LLM non-determinism in hiring is harmful and validates the article's concerns about resume filtering ↑Such inconsistent AI-based filtering likely violates EU anti-discrimination laws ↓From a recruiter's volume perspective, even a flawed 35% pass rate is acceptable given applicant flood ↑The scoring criteria themselves (e.g. requiring open source/projects) unfairly disadvantage many qualified candidates ↓The author misunderstands determinism; reproducibility isn't necessarily desirable here and is technically achievable