Skip to content

AI Candidate Matching and Resume Parsing: How They Work Together

Two AI features solve different parts of the same recruiting problem.

Candidate matching helps you find people already in your database who resemble a strong profile. Resume parsing turns incoming CVs into structured, searchable records. Used together, they reduce manual screening and make the database more useful over time.

What candidate matching does

Candidate matching compares a chosen candidate profile against the rest of your database and returns profiles that look similar. In an AI recruiting tool, this goes beyond keyword overlap: the system weighs a range of attributes such as skills, experience, qualifications, location, language, job titles, and industry. The result is a ranked list of candidates who resemble the original profile, which is useful when you want to replicate a successful placement or build a shortlist around a known strong fit.

How bimetric scoring works

Bimetric scoring compares two candidate profiles in both directions and returns a match rate. It does not reduce a CV to a keyword checklist; it reads the profile as a whole and weighs the attributes recorded there. A higher score means the two profiles line up more closely. The system presents up to 25 matches for the primary candidate, each with a score from 0 to 100. That cap keeps the result set focused, but it also means the list is a shortlist rather than an exhaustive census of the database.

The matching workflow in practice

The flow is deliberately short. Pick the candidate whose profile you want to use as the reference, then click the star icon in the top-right corner of that profile. A panel opens listing candidates whose profiles line up with the reference, and each row carries a match score. From there you can assign candidates to specific roles, add them to hotlists, send emails, propose them to contacts, or take other related actions. The same feature is also reachable from the candidate pipeline tab inside a job profile, which is useful when you start from the role rather than from a person.

What resume parsing does

A resume parser reads a CV and pulls out the details a recruiter would otherwise retype: the candidate's name and contact details, skills, work history, education, social media URLs, and similar fields. Those details land in structured fields rather than staying locked inside an attachment. That structure is what makes a candidate database searchable and filterable. Without parsing, most of that detail cannot be queried reliably. With parsing, each new resume becomes a set of attributes you can match, filter, and enrich.

Parser capabilities worth checking

The parser is integrated with Sovren, a resume-reviewing technology, which the product describes as improving accuracy and depth of extraction. It handles CVs in multiple languages, which matters for agencies sourcing across regions. It accepts resumes from email attachments, job application forms, job boards, and other sources. It can also update empty fields on an existing candidate profile while you edit a candidate or use in-line edit, so an updated CV can refresh a stale record. Finally, it can scan applications or applicant profiles when candidates are added through the Public API or Zapier, which lets the parser sit inside an existing intake workflow rather than only inside the UI.

Which feature do you need?

Use this checklist to decide where to start.

  • You have a strong candidate and want more people like them in your existing database → candidate matching.
  • You have a high volume of incoming CVs and need them searchable → resume parsing.
  • Your database has many incomplete profiles → parsing first, then matching.
  • You are sourcing across languages → confirm the parser's multilingual coverage before relying on it.
  • You add candidates through an API or Zapier → check that parsing runs on that intake path.
  • You need a shortlist fast → matching returns up to 25 scored profiles, not an exhaustive list.
  • You need to enrich an existing record → use the parser's empty-field update path rather than creating a duplicate.

A worked example

Suppose you have just placed a senior data engineer and want to build a bench for the next similar role. You open that candidate's profile and select the star icon. The system returns up to 25 candidates, each with a score between 0 and 100. A profile scoring 92 shares most of the reference candidate's skills, experience, education, location, language, job title, and industry signals. A profile scoring 41 may share a job title but differ on location or seniority. You can then assign the high scorers to the open role, add them to a hotlist, or send a introductory email — all without leaving the matching window. The score is a prioritization aid, not a hiring decision; the recruiter still reviews the profile.

Where these features can fall short

Both features depend on input quality. Matching is only as good as the profile data it compares; a sparse or inconsistent candidate record will produce weaker or misleading scores. Parsing depends on the completeness and clarity of the resume itself, so a poorly formatted or partial CV can yield incomplete fields. The source publisher states that it continually works to handle a wide range of data variations, but the practical lesson is the same: clean intake and complete profiles improve both features. Treat the match score as a ranking signal and verify the underlying profile before acting on it.

Frequently asked questions

Can I try the platform before committing?

The source publisher states that a free trial is available so you can explore the platform and see how it fits your recruitment workflow. Pricing plans are published separately and should be checked for current terms.

Are there limitations to AI candidate matching and resume parsing?

Both features are designed to be efficient and versatile, but their effectiveness depends on the quality of the input data. The completeness and clarity of information in resumes affects parsing precision and matching accuracy. The source publisher states that it continually works to handle a wide range of data variations.

What is bimetric scoring?

It is a scoring method that compares two candidate profiles in both directions and returns a match rate. Rather than testing for a keyword string, it reads the profile as a whole — skills, experience, education, location, language, job titles, industry, and more — to produce a relevance score.