Candidate Matching

Candidate matching is the algorithmic comparison of a candidate's profile against a job's requirements, producing a fit score used to rank applicants — from simple keyword overlap in older systems to weighted multi-dimension semantic scoring in current ones.

In one paragraph

Matching is always relative to one job: the same candidate legitimately scores 85 against one posting and 40 against another. First-generation matching counted shared keywords, which rewarded resume-stuffing and missed synonyms; modern matching scores meaning across dimensions (skills, experience, stack, seniority, stability) with explicit weights and, in auditable systems, a written reason per score. The SharpAPI implementation returns 20 such dimensions on a 0–100 scale.

Example

skills_match: 80 — "Strong PHP/MySQL; Laravel not mentioned explicitly" — a score plus its evidence, which is what separates rankable from defensible.

Go deeper

Semantic vs keyword matching · what counts as a good score · resume screening, the decision layer that consumes match scores.