The AI Matcher helps recruiters quickly see how well a candidate's CV fits a job description. It gives every application a match score, so you can prioritize your review instead of reading every CV top to bottom.
What the AI Matcher does
When a candidate applies to a job, the AI Matcher compares the content of their CV to the job description and calculates a similarity score from 0 to 100. The higher the score, the more the CV's content overlaps with what the job description is asking for.
The score is a support signal, not a decision. It's there to help you triage applications faster. It doesn't rank, filter, or remove any candidates on its own, and it never replaces the hiring team's judgment. You choose whether and how to use the score, including whether to sort your candidate list by it.
Turning it on
AI Matching isn't on by default. It's activated by Jobylon on request, and once it's available on your account, a permitted user or Manager can turn it on per job or recruitment, it's not an account-wide setting. This means you decide which recruitments use it.
Once it's enabled for a job, the match score appears next to each applicant in the candidate list, alongside a tooltip or disclaimer reminding you that it's AI-generated and meant as a guideline, not a final verdict.
How the score is calculated
The AI Matcher uses a pre-trained, open-source language model that reads text and turns it into a numerical representation of its meaning. This lets it compare a CV and a job description based on the substance of what's written, not just keyword overlap.
A few things worth knowing about the model:
It's static. It doesn't learn or adapt based on your data over time, and it isn't retrained on customer applicant data.
It only looks at the text content of the CV and job description. It doesn't use demographic attributes such as age, gender, or ethnicity as input.
It was fine-tuned using curated, synthetically generated job ad and candidate data, in the languages Jobylon supports.
Candidate consent
You can optionally add a disclaimer to the application form so candidates know AI is used as part of the process, and let them opt out of being matched by AI when they apply. If a candidate opts out, their application is not scored.
Human oversight
A person is always in the loop. The AI Matcher surfaces a score, recruiters and hiring managers decide what to do with it, and every match can be disregarded or overridden. Jobylon is designed this way deliberately, both because it produces better hiring decisions and because it's core to how we think about responsible use of AI in recruitment.
Data and privacy
No customer-specific applicant data is used to train the AI Matcher. The underlying model is a general-purpose, pre-trained embedding model, adapted using synthetic and manually curated data rather than any customer's real candidate data.
Want more detail?
This article covers the AI Matcher at a high level. If you need deeper technical documentation, for example system architecture, data flow, or our AI Act risk assessment, we're happy to share that on request. Just reach out to your Jobylon contact.