Autism-Friendly Job Description Adapter

What it does

Takes an existing job description/ad and rewrites it sentence by sentence to lower application barriers for autistic and broader neurodivergent candidates — without diluting the role's essential functions. Two output layers: (1) a copy-paste-ready adapted ad, and (2) a line-by-line audit + 0-100 score + readability band showing what changed and why.

The core move is translation, not softening: vague, performative, similarity-demanding phrasing is reduced to observable behavior/output. "Excellent communication" becomes "convey requests clearly and on time, in writing or speech"; "multitasking under stress" becomes (only if genuinely required) "prioritize 2-3 concurrent tasks"; "culture fit" is usually removed entirely as a social-similarity filter. Research confirms this: long must-have lists and coded language deter neurodivergent applicants.

When to use it

When you have a real job description / ad / interview-invite for a role and need it accessible, clear, and fair for neurodivergent candidates. For HR teams, inclusive-hiring consultants, diversity/accessibility leads, and hiring managers who want to audit ad language while preserving the role's true core — including the interview/application process, where most barriers actually live.

Method / frameworks

• Essential vs Marginal Functions (ADA Title I job analysis): is each task the role's core or an incidental/historical requirement? Marginal is removed; essential is written concretely and measurably.
• W3C COGA + WCAG 2.2 (SC 3.1 Readable): short sentences, one idea each, concrete words, expanded acronyms, numbered steps.
• Coded/ableist language scan — the Gaucher, Friesen & Kay (2011) gendered-wording method extended to neurodiversity; performative/exclusionary patterns are converted to behavior-based equivalents.
• JAN SOAR (Job Accommodation Network): a low/zero-cost concrete accommodation per barrier (questions shared 24h ahead, quiet space, written instructions).
• Universal Design (Mace, NC State 7 Principles) + UDL (CAST): clarity yields a better application experience for everyone.
• Readability metrics — Ateşman and Bezirci-Yılmaz for Turkish (Ateşman is adapted from Flesch-Kincaid, based on word/sentence length); Flesch-Kincaid for English. Simplification is measured, not guessed.
• TR legal frame: Labor Law 4857 (equal treatment), Law 5378 (reasonable accommodation), KVKK (health data is special-category — asking/recording a diagnosis is prohibited).