Anti AI-Slop ATS System
Auto-apply tools and LLM-filled forms flood hiring pipelines with “AI slop”:
About the project
Problem: Auto-apply tools and LLM-filled forms flood hiring pipelines with “AI slop”: applications that look real but lack genuine intent. Standard ATS data (résumés and text alone) can’t expose bot speed, paste bursts, or repeat fingerprints, so recruiters waste time on noise.
Who it helps: Recruiters and hiring managers who need an explainable likely bot vs likely human signal before scheduling interviews, and teams that run honeytrap career pages to study auto-apply behavior.
Solution: Anti-Slop ATS is a first-party careers site plugin. The apply flow captures field telemetry and browser fingerprint signals, then scores submissions with P(auto), confidence, and transparent feature reasoning via a deterministic model, not a black-box LLM. Recruiters can see a "likely human or likely auto-apply" classification for the applications.
Impact: Faster triage, less slop in the review queue, and a practical path to test real auto-apply tools, without auto-rejecting candidates or replacing a full ATS.