How It Works

From job requirements to ranked shortlist in minutes

Candidharbor applies your criteria systematically to every application. No training data setup, no ML team, no mysterious scores.

Connect your ATS

Candidharbor integrates directly with your applicant tracking system. New applications flow in automatically as candidates apply. Setup takes under 15 minutes with our guided connector wizard.

Supported systems include Meridian ATS, Clarity Recruit, Vantage Hire, and Nexus HRIS. Don't see yours? Our REST API connects to any ATS with webhook support.

Your ATS
Candidharbor
Ranked Shortlist

Define your requirements

Paste your job description or manually enter your must-have and nice-to-have criteria. Candidharbor structures them into a weighted scoring rubric and shows you exactly what it will evaluate.

You can adjust weights before screening starts: years of experience might count more than a specific tool, or you can require certain credentials as hard filters that eliminate candidates before scoring begins.

Every application is screened

Candidharbor reads each resume and cover letter, extracts relevant signals, and scores each candidate against your rubric. A 400-application pool is processed in under 8 minutes.

The evaluation focuses on experience, skills, and demonstrated outcomes, not name, school, or address. Every candidate is measured against the same bar.

SOC 2 Type II compliant data handling

Review your ranked shortlist

Results land in your ATS as a ranked list. Each candidate has a score and a plain-English summary: which requirements they met, which they didn't, and where the system had low confidence.

You stay in control. The shortlist is a starting point, not a final decision. You decide who to advance and why, with the benefit of systematic context you didn't have before.

What Makes It Work

The key decisions we made in the design

Rubric-first, not model-first

We don't train a general hiring model and hope it generalizes. Your rubric defines the evaluation. The AI extracts evidence from resumes; your criteria decide what that evidence means.

Traceable scoring

Every score links to a specific piece of evidence in the resume. Click any score to see exactly what the system read and how it interpreted it. No opaque numbers.

Confidence signals

When the system can't confidently assess a requirement from the available text, it says so. Low-confidence flags tell you where to look more carefully yourself.

Rescoring on requirement changes

Change a weight or add a requirement after screening starts. The entire pool is rescored instantly. No need to re-run from scratch.

Run your first shortlist today

Connect your ATS, paste a job description, and see your top 10 candidates in under 10 minutes.