AI candidate matching software
Every applicant matched to your vacancy. Ranked. Explained.
Talent Atrium evaluates every applicant against your vacancy requirements across five structured dimensions. Not keyword counting. Not filtering. A match score, a ranked list, and a written explanation for each candidate before you open a single CV.
Why most AI candidate matching falls short of the promise
AI candidate matching has become a broad label. It covers everything from basic keyword density scoring to genuine semantic evaluation of candidate profiles. The gap between these approaches is significant, and the results reflect it.
Keyword-based matching rewards candidates who optimise their CVs for the words in the job description and penalises those who describe their experience in their own language. It treats a mention of "project management" the same whether the candidate ran a 15-person programme for two years or used the phrase once in a summary. The score reflects the text, not the substance.
Genuine AI matching evaluates the underlying alignment: is the experience relevant and substantial? Do the skills appear in a context that supports real proficiency? Does the candidate's behavioural profile suit the demands of this specific role? Does the practical situation align? Answering these questions produces a match score you can trust. Counting keywords produces a score that rewards the wrong things.
What AI matching needs to get right
Matching must be against the specific vacancy, not a general template
A match score that uses a generic skills library or industry-average profile tells you how similar a candidate is to a hypothetical archetype. It does not tell you how well they fit your specific vacancy with your specific requirements. Meaningful matching requires the requirements to be extracted from what you actually need.
Context determines relevance, not term frequency
A candidate who spent three years as a senior analyst has different experience from one who spent three years in a junior administrative role, even if both CVs contain the same keywords. AI matching needs to assess the context of experience and skills, not just their presence.
Behavioural fit is part of the match, not a separate question
Technical alignment is necessary but not sufficient. A candidate who meets the skill requirements but has a behavioural profile that is poorly suited to the working environment creates a different kind of hire risk. AI matching should incorporate behavioural signals into the overall evaluation, not treat them as optional extras.
The output must be explainable
A match score without reasoning is difficult to act on. If you cannot see why a candidate received the score they did, you cannot apply your own judgement to override or confirm the ranking. The output of AI matching should always include written reasoning that makes the evaluation transparent.
How Talent Atrium matches candidates to vacancies
The matching process is structured, automatic, and vacancy-specific. Every evaluation runs against the requirements extracted from your job description, not against a generic template.
Requirements are extracted from your vacancy
Post your job description and the platform extracts the experience profile, skills, qualifications, behavioural requirements, and practical fit factors. You review and adjust before publishing. These requirements are what every candidate is matched against.
Every applicant is evaluated automatically at submission
When a candidate applies — submitting their CV, psychometric profile, and demographic details — the AI evaluates their application across all five dimensions immediately. No queue. No batch processing at the end of the day.
A match score and written rationale are generated
Each candidate receives an overall match score from 0 to 100 and a written compatibility report. The report explains how they performed on each dimension, what their strongest alignment points are, and where the gaps are relative to your requirements.
You work from a ranked list
Candidates are ordered by match score in the recruiter dashboard. The highest-scoring candidates appear at the top. You review the written reports for the candidates you want to progress and shortlist accordingly.
Five dimensions. One match score.
The matching framework covers every aspect of candidate fit. Each dimension produces a sub-score. The five sub-scores combine into a single match score from 0 to 100. The weighting reflects what your specific vacancy requires.
Experience
Is the candidate's prior work history relevant and substantial enough for this role? Depth and recency both factor in, not just the presence of similar job titles.
Skills
Technical and practical capabilities evaluated against your vacancy. The AI assesses whether the work history supports genuine proficiency, not just whether the skill is mentioned.
Qualifications
Educational background and professional certifications evaluated against your stated requirements. The match reflects actual alignment, including cases where a candidate is over- or under-qualified.
Behavioural Fit
Workplace behaviour patterns from the candidate profile matched against the demands of the role. This dimension captures how a candidate works, not just what they have done.
Role Alignment
Salary expectations, location, availability, and practical factors. A candidate who cannot take the role is matched accordingly, regardless of their substantive qualifications.
What the AI matching output gives you
The output is designed for fast, confident decision-making. Every element of the matching result is structured to reduce the time between application arrival and shortlist decision.
Match score from 0 to 100
A single figure representing the overall alignment between the candidate and your vacancy requirements across all five dimensions.
Dimension sub-scores
Individual scores for each of the five dimensions, showing exactly where the match is strong and where there are gaps.
Written compatibility report
A plain-language explanation of the match. Not a summary of the CV — a structured evaluation of how the candidate aligns with your specific requirements.
Ranked candidate list
Every applicant ordered by match score. Work from the top of the list rather than sorting through an unranked pile.
Structured feedback for non-shortlisted candidates
When you close the vacancy, rejected applicants receive feedback derived from their match evaluation automatically.
Related tools and resources
Candidate Misalignment Detector
Find the gap between what your vacancy says and what candidates hear. Free tool.
Rubric Builder
Build the evaluation criteria that the AI matches candidates against. Free tool.
Candidate screening software
Every applicant evaluated before you open a single CV. Ranked shortlist included.
Explainable hiring decisions
Shortlists built on written reasoning you can show to any stakeholder.
Candidate screening software vs ATS
What each system does and where AI matching fits into the picture.
Frequently asked questions
What is AI candidate matching?
AI candidate matching is the automated evaluation of job applicants against the specific requirements of a vacancy. The best AI matching systems assess the substantive alignment between a candidate's experience, skills, qualifications, behavioural profile, and practical fit factors — not just keyword presence. The output is a ranked list with documented reasoning for each evaluation.
How is AI matching different from keyword-based screening?
Keyword-based screening identifies whether specific terms appear in a CV. AI matching evaluates the context behind the words — whether experience is relevant and deep, whether skills appear in a context that supports genuine proficiency. The score reflects alignment, not term frequency.
What does Talent Atrium match candidates against?
Talent Atrium extracts structured requirements from your job description when you post a vacancy. Every candidate is evaluated against these specific requirements across five dimensions: experience, skills, qualifications, behavioural fit, and role alignment.
Can I trust an AI match score to make a hiring decision?
The match score is a starting point, not a final decision. It tells you which candidates have the strongest documented alignment with your requirements. The written compatibility report explains the reasoning. You use this to prioritise who to review — the hiring decision remains with you.
Does the AI evaluate every applicant or just a subset?
Every applicant is evaluated. There is no pre-filtering step that removes candidates before the AI assessment runs. Talent Atrium evaluates every application against the full five-dimension framework and ranks the complete pool.
How does the AI handle candidates who do not fit the standard profile?
The evaluation is based on your vacancy requirements, not a standard profile. A candidate who took an unconventional route will be assessed on their actual experience and skills. The compatibility report will reflect where they align and where they fall short.
What output does the AI matching produce?
For each candidate: an overall match score from 0 to 100, a breakdown across the five evaluation dimensions, and a written compatibility report in plain language. Candidates are ranked from highest to lowest match score.
Does AI matching work for all types of roles?
Yes. The evaluation framework adapts to the requirements extracted from your specific vacancy. Whether you are hiring for a technical, leadership, or customer-facing role, the five-dimension framework applies — weighted towards what your vacancy actually requires.
AI candidate matching
Match every applicant to your vacancy automatically
Post a vacancy and every applicant is matched against your requirements across five structured dimensions. A ranked list and written compatibility report for each candidate — before you open a single CV.