GitHire: An Evidence-Based, Explainable Candidate Evaluation Platform Combining Resume Analysis and Public Repository Signals
Year: 2026
Type: preprint
Abstract: Resume screening and applicant tracking systems (ATS) favour candidates whose resumes match keywords, and can miss developers whose ability is visible in their public projects. We present GitHire, a hiring platform that evaluates candidates from three kinds of evidence: the parsed resume, public GitHub repositories and self-declared projects. It produces three deterministic, explainable outputs: an ATS check (0–100), a Job Ready Score with a stated confidence level, and a candidate–job match score whose every point is itemised. Missing evidence is left out of a score instead of being counted as zero, repository popularity (stars, forks) is never used, and recruiters see scores and explanations rather than resume contents. We describe the architecture (a modular monolith with a React front end and a FastAPI back end), the scoring rules, and design choices made with regulation of AI in hiring in mind. We report no empirical evaluation yet and set out an evaluation plan with a bias audit.
Language: English
Topic: AI and HR Technologies
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7236754731
