Automate Resume Screening With Two API Calls

A resume screening API automates the first review pass over an applicant pool: extract each candidate's data, score it against the job description, and return a ranked shortlist with reasons. SharpAPI implements this as two composable endpoints — parse_resume and resume_job_match_score.

The screening loop, end to end

The pattern is the same whether the pool holds 20 applicants or 2,000, and because both endpoints are async, the whole pool runs in parallel:

for each application in pool:
    job = POST /hr/resume_job_match_score          # file + job description
    scores[application] = poll(job.status_url)      # 20 dimensions + reasons

shortlist = sort(scores, by="overall_match", desc)  # recruiters start at the top
review(shortlist[:10].explanations)                 # written evidence per score

Need the structured profile too — for the candidate record, search, or dedupe? Run parse_resume on the same file and store the 50+ field JSON alongside the scores.

What separates screening from keyword filtering

Keyword filters reject the developer who wrote “built REST services in Django” for a “Python backend” role. Semantic scoring reads meaning: the match-score endpoint evaluates 20 dimensions of fit, weights must-haves at 3× nice-to-haves, and explains each headline number in writing. The difference is measured in good candidates that stop falling through — more in semantic vs keyword matching.

Screening you can put in front of a regulator

Automated screening is exactly what NYC Local Law 144 and the EU AI Act regulate. Two properties of this pipeline do the heavy lifting: PII never reaches the scoring model — names, ages, nationalities and photos are stripped after parsing — and every score ships with a written explanation. Ranked, and defensible. The compliance details live in our AI screening law guide.

Sizing the workload